Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 января 2011 г., 22:57:17:

> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.
I (as customer with server which has problem) could ask techsupport
tomorrow. But, maybe, they didn't answer, because I already have
answer that FreeBSD is not supported :(

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 января 2011 г., 22:57:17:

> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.
I (as customer with server which has problem) could ask techsupport
tomorrow. But, maybe, they didn't answer, because I already have
answer that FreeBSD is not supported :(

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Pyun.
You wrote 11 ������ 2011 �., 23:00:07:

> rgephy(4) currently always use auto-negotiation to work-around link
> establishment issues reported in past.
I think, it is the root of the problem. Autonegotiation is DISABLED on
these ports. I think, some additional mediaopt (like
force-half-duplex) for rgephy(4) will be solution.

> For your case, the only way to address the issue at this moment is
> to use auto-negotiation but that would establish 1000baseT link
> which would add cost for you. Alternatively request half-duplex
> configuration to the provider to get a agreed link duplex.
Maybe, adding new mediaopt is not very hard? Or is it?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 января 2011 г., 22:57:17:

> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.
I (as customer with server which has problem) could ask techsupport
tomorrow. But, maybe, they didn't answer, because I already have
answer that FreeBSD is not supported :(

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Pyun.
You wrote 11 ������ 2011 �., 23:00:07:

> rgephy(4) currently always use auto-negotiation to work-around link
> establishment issues reported in past.
I think, it is the root of the problem. Autonegotiation is DISABLED on
these ports. I think, some additional mediaopt (like
force-half-duplex) for rgephy(4) will be solution.

> For your case, the only way to address the issue at this moment is
> to use auto-negotiation but that would establish 1000baseT link
> which would add cost for you. Alternatively request half-duplex
> configuration to the provider to get a agreed link duplex.
Maybe, adding new mediaopt is not very hard? Or is it?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 10:37:31PM +0300, Lev Serebryakov wrote:
> Hello, Brian.
> You wrote 11 ?????? 2011 ?., 22:29:13:
>
> > basic mode: 100 Mbit, full duplex
> > link partner: 100baseTx-HD
> It looks VERY strange. How could id be?

This is normal if the link partner doesn't do autonegotiation.
The system has to assume a simple repeater hub and do half duplex.
If on the other hand the switch is hard set to full-duplex then they
can't expect autonegotiation to do it right.
However hard setting under FreeBSD as well should work.

--
B.Walter <> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 января 2011 г., 22:57:17:

> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.
I (as customer with server which has problem) could ask techsupport
tomorrow. But, maybe, they didn't answer, because I already have
answer that FreeBSD is not supported :(

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Pyun.
You wrote 11 ������ 2011 �., 23:00:07:

> rgephy(4) currently always use auto-negotiation to work-around link
> establishment issues reported in past.
I think, it is the root of the problem. Autonegotiation is DISABLED on
these ports. I think, some additional mediaopt (like
force-half-duplex) for rgephy(4) will be solution.

> For your case, the only way to address the issue at this moment is
> to use auto-negotiation but that would establish 1000baseT link
> which would add cost for you. Alternatively request half-duplex
> configuration to the provider to get a agreed link duplex.
Maybe, adding new mediaopt is not very hard? Or is it?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 10:37:31PM +0300, Lev Serebryakov wrote:
> Hello, Brian.
> You wrote 11 ?????? 2011 ?., 22:29:13:
>
> > basic mode: 100 Mbit, full duplex
> > link partner: 100baseTx-HD
> It looks VERY strange. How could id be?

This is normal if the link partner doesn't do autonegotiation.
The system has to assume a simple repeater hub and do half duplex.
If on the other hand the switch is hard set to full-duplex then they
can't expect autonegotiation to do it right.
However hard setting under FreeBSD as well should work.

--
B.Walter <> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 10:50:49PM +0300, Lev Serebryakov wrote:
> Hello, Artyom.
> You wrote 11 ?????? 2011 ?., 22:39:33:
>
> >> link partner: 100baseTx-HD
> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> > Looks very strange for me... 'HD' means half-duplex?
> Yep, I've noticed that too...
>
> > May be linux driver defaults to full-duplex if autoneg fails?..
> Or disabled... And it works -- very strange. And FreeBSD uses
> half-duplex even with given "media-opt" and network is dramatically
> slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
> installer).
>
I'm not surprised that it doesn't work with autonegotian if autonegotian
is disabled.
If Linux does full-duplex without autonegotiation then _they_ do it wrong
and Hetzner shouldn't rely on wrong behavour.
However since you seem to have access it might be interesting to know the
exact network device and debug why hard setting is broken.

--
B.Walter <> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.

Hello, Freebsd-net.

Very large and famous (due to very attractive prices) hosting
provider Hetzner.de discards FreeBSD support on dedicated servers,
because these servers can niot negotiate 100Mbit/DUPLEX when
switches' ports are limited to 100Mbit (1Gbit connection costs
additional money) only under FreeBSD. Linux works fine.

Switches known to be Juniper e3k series.

MoBos of servers are different assortment of MSI MoBos with Realtek
(re driver) network-on-board.

Symptjms are: NIC can not negotiate/set duplex when switch port is
limited to 100Mbit/Duplex. Duplex can not be set even manually via
"ifconfig":


media: Ethernet 100baseTX (100baseTX )

Is it know problem? Maybe, -CURRENT driver has fix for it?

Unfortunately, I can not provide more information, as I don't have
server at Hetzner (I'm planning to order one, but due to these
problems, I'm not sure now, as I need FreeBSD), and all this
information is collected in communication with people who HAVE servers
with FreeBSD installed.

Again, I know, that Realtek NICs are crap, but "everybody says" that
Linux doesn't have THIS problem with THESE boards and switches.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Yamagi.
You wrote 11 ������ 2011 �., 13:30:23:

> Hi,
> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.
It is known, that problems are in DC 13 and everything wotrks fine
in DC 11 and DC 12.

I've discussed this problem in local (Russian-speaking) FreeBSD
community, and there are several people in DC 13 who HAVE these
problems and found different solutions, but all non-technical ones:
order gigabit connectivity, or pay for moving servers to other (old)
DCs...

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Yamagi Burmeister wrote:

>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
>>
>> Switches known to be Juniper e3k series.

...

> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> just fine. I've never seen this strange negotiation problem myself. But
> maybe I was just lucky and got working mainboard and nic combinations.
> So if further information is needed, I'm happy to provide it.

A lot of us do. There is a problem with the re(4) setup as well in
that if you do not send packets out yourself the port takes a very
long time to come up and unblocked. I haven't discussed that with
them or tested with an updated HEAD (since end of October).

But yes, I am running HEAD on an EQ4 as well. If you have problems
and a personal email contact at Hetzner feel free to talk to me.
I am "local" (a couple of 100km away in the same country) and a FreeBSD
committer and I can probably figure things out with them or properly
proxy requests.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Bjoern A. Zeeb wrote:

>> I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
>> just fine. I've never seen this strange negotiation problem myself. But
>> maybe I was just lucky and got working mainboard and nic combinations.
>> So if further information is needed, I'm happy to provide it.

> A lot of us do. There is a problem with the re(4) setup as well in
> that if you do not send packets out yourself the port takes a very
> long time to come up and unblocked. I haven't discussed that with
> them or tested with an updated HEAD (since end of October).

I never said that this problems doesn't exists. :) Lev Serebryakov said
that everythings works fine in DC11 and DC12, my servers are in DC12. so
I was just lucky...

> But yes, I am running HEAD on an EQ4 as well. If you have problems
> and a personal email contact at Hetzner feel free to talk to me.
> I am "local" (a couple of 100km away in the same country) and a FreeBSD
> committer and I can probably figure things out with them or properly
> proxy requests.

Sadly no. My only contact to Hetzner is the service e-mail adress and
the phone number for business clients. They are for all customers and
probably can't help with such problems. There are special technical
contacts for each DC, but those are only available for customers with
hardware in that DC and with specific problems. So someone with a server
in DC13 could write a service request in which the problem is explained
and ask for help. Maybe they're willing ton assistent in tracking down
and solving the problem.

Ciao,
Yamagi

--
Homepage: www.yamagi.org
Jabber:
GnuPG/GPG: 0xEFBCCBCB
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.

How are the switches being forced to 100/full?

If they're doing so by disabling autonegotiation, then that's where
some grief may come from.

If it's not, then ignore the rest of this email. :)

For certain hardware combos, I've seen even Linux servers (on Dell
hardware) fail to autonegotiate properly.

Here's the set of litany I trot out when I have to deal with
customer's issues surrounding gigabit and autonegotiation:

-------------

With the advent of 1000T networking, the specs says that autonegotation
needs to be enabled:

http://etherealmind.com/2008/07/15/ethernet-autonegotiation-works-why-how-standard-should-be-set/

" A major problem is that many people are also hard setting Gigabit
Ethernet, and this is causing major problems. Gigabit Ethernet
must have auto-negotiation ENABLED to allow negotiation of master
/ slave PHY relationship for clocking at the physical layer.
Without negotiation the line clock will not establish correctly
and physical layers problems can result."

Further, this doc from Dell:

http://www.dell.com/content/topics/global.aspx/power/en/ps1q01_hernan?c=us&cs=555&l=en&s=biz

Cites:

"In addition, the 1999 standard for Gigabit over copper cabling,
IEEE Std 802.3ab, added the following enhancements to the
Auto-Negotiation standard:"

* Mandatory auto-negotiation for 1000BaseT
* Configure master and slave modes for the PHY

Further:

http://en.wikipedia.org/wiki/Autonegotiation

"The debatable portions of the autonegotiation specifications
were eliminated by the 1998 version of IEEE 802.3. In 1999, the
negotiation protocol was significantly extended by IEEE 802.3ab,
which specified the protocol for gigabit Ethernet, making
autonegotiation mandatory for 1000BASE-T gigabit Ethernet over
copper."

> _______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 19:38:25:


>> Very large and famous (due to very attractive prices) hosting
>> provider Hetzner.de discards FreeBSD support on dedicated servers,
>> because these servers can niot negotiate 100Mbit/DUPLEX when
>> switches' ports are limited to 100Mbit (1Gbit connection costs
>> additional money) only under FreeBSD. Linux works fine.
> How are the switches being forced to 100/full?
I don't know, I never work with Juniper e3k switches (And any other
Juniper products at all).

All I know, that older Juniper Switches in not-so-new DCs of same
provider doesn't have this problem, and, on other hand, Linux and
Windows 2008 don't have problems with new ones too.

> If they're doing so by disabling autonegotiation, then that's where
> some grief may come from.
Linux work with autonegotiation, as I can see (It is outpuit from
Rescue Linux system on SAME my server, where FreeBSD shows
half-duplex even if forced to full-duplex):

root@rescue ~ # mii-tool -v eth0
eth0: 100 Mbit, full duplex, link ok
product info: vendor 00:07:32, model 17 rev 2
basic mode: 100 Mbit, full duplex
basic status: link ok
capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 100baseTx-HD
root@rescue ~ # ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: No
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: off
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000033 (51)
Link detected: yes
root@rescue ~ #

So, it seems, that autonegotiation is disabled, but it works for
Linux, and manual setting of media and mediaopt doesn't help FreeBSD.

Also, please note, that when port is in 1Gib mode (which can be buyed
for additional money, which I can not afford) FreeBSD works fine.

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 01:53:30PM +0300, Lev Serebryakov wrote:
> Hello, Yamagi.
> You wrote 11 ?????? 2011 ?., 13:30:23:
>
> > Hi,
> > I've got several Hetzner EQ4 and on all these machines FreeBSD 8.1 runs
> > just fine. I've never seen this strange negotiation problem myself. But
> > maybe I was just lucky and got working mainboard and nic combinations.
> > So if further information is needed, I'm happy to provide it.
> It is known, that problems are in DC 13 and everything wotrks fine
> in DC 11 and DC 12.
>
> I've discussed this problem in local (Russian-speaking) FreeBSD
> community, and there are several people in DC 13 who HAVE these
> problems and found different solutions, but all non-technical ones:
> order gigabit connectivity, or pay for moving servers to other (old)
> DCs...
>

If the latter means that the servers are physically moved as opposed
to a new one being allocated this implies that re(4)/rgephy(4) isn't
the sole factor responsible for this problem. In any case it would be
helpful to have the corresponding dmesg bits as the Linux counterpart
does some black magic for certain hardware versions when setting the
media manually but re(4) doesn't which might be relevant in this
scenario.

Marius

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Brian.
You wrote 11 ������ 2011 �., 22:29:13:

> basic mode: 100 Mbit, full duplex
> link partner: 100baseTx-HD
It looks VERY strange. How could id be?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Marius.
You wrote 11 ������ 2011 �., 22:36:44:


>> I've discussed this problem in local (Russian-speaking) FreeBSD
>> community, and there are several people in DC 13 who HAVE these
>> problems and found different solutions, but all non-technical ones:
>> order gigabit connectivity, or pay for moving servers to other (old)
>> DCs...
> If the latter means that the servers are physically moved as opposed
> to a new one being allocated this implies that re(4)/rgephy(4) isn't
> the sole factor responsible for this problem. In any case it would be
It is known, that DC 13 has new Juniper e3k switches, and older DCs
have older Juniper equipment.

> helpful to have the corresponding dmesg bits as the Linux counterpart
> does some black magic for certain hardware versions when setting the
> media manually but re(4) doesn't which might be relevant in this
> scenario.
Here it is from Rescue FreeBSD system:

re0: port 0xe800-0xe8ff mem 0xfbeff000-0xfbefffff,0xf8ff0000-0xf8ffffff irq 16 at device 0.0 on pci6
re0: Using 1 MSI messages
re0: Chip rev. 0x3c000000
re0: MAC rev. 0x00400000
miibus0: on re0
rgephy0: PHY 1 on miibus0
rgephy0: 10baseT, 10baseT-FDX, 10baseT-FDX-flow, 100baseTX, 100baseTX-FDX, 100baseTX-FDX-flow, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, 1000baseT-FDX-flow, 1000baseT-FDX-flow-master, auto, auto-flow
re0: Ethernet address: 6c:62:6d:a7:bb:37
re0: [FILTER]


And, please note very strange output from Linux's mii-tool and
ethtool in previous my message to list: mismatch between "basic mode"
and "link partner".

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:29, Lev Serebryakov ?????:
>>
>> root@rescue ~ # mii-tool -v eth0
>> eth0: 100 Mbit, full duplex, link ok
>> product info: vendor 00:07:32, model 17 rev 2
>> basic mode: 100 Mbit, full duplex
>> basic status: link ok
>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>> 10baseT-FD 10baseT-HD
>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>> flow-control
>
>
>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Looks very strange for me... 'HD' means half-duplex?
>
> May be linux driver defaults to full-duplex if autoneg fails?..

That's probably just because it cannot tell what the peer really uses
(autoneg disabled) and prints the sane fall-back but I don't know the code.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 ������ 2011 �., 22:39:33:

>> link partner: 100baseTx-HD
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Looks very strange for me... 'HD' means half-duplex?
Yep, I've noticed that too...

> May be linux driver defaults to full-duplex if autoneg fails?..
Or disabled... And it works -- very strange. And FreeBSD uses
half-duplex even with given "media-opt" and network is dramatically
slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
installer).

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. 11.01.2011 21:48, Bjoern A. Zeeb пишет:
> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>
>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>
>>> root@rescue ~ # mii-tool -v eth0
>>> eth0: 100 Mbit, full duplex, link ok
>>> product info: vendor 00:07:32, model 17 rev 2
>>> basic mode: 100 Mbit, full duplex
>>> basic status: link ok
>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>
>>
>>> link partner: 100baseTx-HD
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>> Looks very strange for me... 'HD' means half-duplex?
>>
>> May be linux driver defaults to full-duplex if autoneg fails?..
>
> That's probably just because it cannot tell what the peer really uses
> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>

Is it possible to see status from corresponding port on Juniper switch?
Config part for this port on the switch would be also very interesting.

--
Sincerely yours,
Artyom Viklenko.
-------------------------------------------------------
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 12:47:29PM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-net.
>
> Very large and famous (due to very attractive prices) hosting
> provider Hetzner.de discards FreeBSD support on dedicated servers,
> because these servers can niot negotiate 100Mbit/DUPLEX when
> switches' ports are limited to 100Mbit (1Gbit connection costs
> additional money) only under FreeBSD. Linux works fine.
>
> Switches known to be Juniper e3k series.
>
> MoBos of servers are different assortment of MSI MoBos with Realtek
> (re driver) network-on-board.
>
> Symptjms are: NIC can not negotiate/set duplex when switch port is
> limited to 100Mbit/Duplex. Duplex can not be set even manually via
> "ifconfig":
>
>
> media: Ethernet 100baseTX (100baseTX )
>
> Is it know problem? Maybe, -CURRENT driver has fix for it?
>
> Unfortunately, I can not provide more information, as I don't have
> server at Hetzner (I'm planning to order one, but due to these
> problems, I'm not sure now, as I need FreeBSD), and all this
> information is collected in communication with people who HAVE servers
> with FreeBSD installed.
>
> Again, I know, that Realtek NICs are crap, but "everybody says" that
> Linux doesn't have THIS problem with THESE boards and switches.
>

I can see what's going on here. Link partner used forced media
configuration, probably 100baseTX/full-duplex, and re(4)'s
resolved link is 100baseTX/half-duplex.
rgephy(4) currently always use auto-negotiation to work-around link
establishment issues reported in past. I don't know how Linux
managed to address link establishment issues for
non-autonegotiation case though. Perhaps a lot of vendor supplied
DSP fixups addressed that issue but I'm not sure.
For your case, the only way to address the issue at this moment is
to use auto-negotiation but that would establish 1000baseT link
which would add cost for you. Alternatively request half-duplex
configuration to the provider to get a agreed link duplex.

See
http://lists.freebsd.org/pipermail/freebsd-amd64/2011-January/013589.html
for details on parallel detection.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, 11 Jan 2011, Artyom Viklenko wrote:

> 11.01.2011 21:48, Bjoern A. Zeeb :
>> On Tue, 11 Jan 2011, Artyom Viklenko wrote:
>>
>>> 11.01.2011 21:29, Lev Serebryakov ?????:
>>>>
>>>> root@rescue ~ # mii-tool -v eth0
>>>> eth0: 100 Mbit, full duplex, link ok
>>>> product info: vendor 00:07:32, model 17 rev 2
>>>> basic mode: 100 Mbit, full duplex
>>>> basic status: link ok
>>>> capabilities: 1000baseT-HD 1000baseT-FD 100baseTx-FD 100baseTx-HD
>>>> 10baseT-FD 10baseT-HD
>>>> advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
>>>
>>>
>>>> link partner: 100baseTx-HD
>>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>
>>> Looks very strange for me... 'HD' means half-duplex?
>>>
>>> May be linux driver defaults to full-duplex if autoneg fails?..
>>
>> That's probably just because it cannot tell what the peer really uses
>> (autoneg disabled) and prints the sane fall-back but I don't know the code.
>>
>
> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.

No but I think we are not getting anywhere; I contacted them this
afternoon and will see if they'll get back to me.

/bz

--
Bjoern A. Zeeb You have to have visions!
Going to jail sucks -- All my daemons like it!
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/jails.html
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Artyom.
You wrote 11 января 2011 г., 22:57:17:

> Is it possible to see status from corresponding port on Juniper switch?
> Config part for this port on the switch would be also very interesting.
I (as customer with server which has problem) could ask techsupport
tomorrow. But, maybe, they didn't answer, because I already have
answer that FreeBSD is not supported :(

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. Hello, Pyun.
You wrote 11 ������ 2011 �., 23:00:07:

> rgephy(4) currently always use auto-negotiation to work-around link
> establishment issues reported in past.
I think, it is the root of the problem. Autonegotiation is DISABLED on
these ports. I think, some additional mediaopt (like
force-half-duplex) for rgephy(4) will be solution.

> For your case, the only way to address the issue at this moment is
> to use auto-negotiation but that would establish 1000baseT link
> which would add cost for you. Alternatively request half-duplex
> configuration to the provider to get a agreed link duplex.
Maybe, adding new mediaopt is not very hard? Or is it?

--
// Black Lion AKA Lev Serebryakov <>

_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 10:37:31PM +0300, Lev Serebryakov wrote:
> Hello, Brian.
> You wrote 11 ?????? 2011 ?., 22:29:13:
>
> > basic mode: 100 Mbit, full duplex
> > link partner: 100baseTx-HD
> It looks VERY strange. How could id be?

This is normal if the link partner doesn't do autonegotiation.
The system has to assume a simple repeater hub and do half duplex.
If on the other hand the switch is hard set to full-duplex then they
can't expect autonegotiation to do it right.
However hard setting under FreeBSD as well should work.

--
B.Walter <> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Tue, Jan 11, 2011 at 10:50:49PM +0300, Lev Serebryakov wrote:
> Hello, Artyom.
> You wrote 11 ?????? 2011 ?., 22:39:33:
>
> >> link partner: 100baseTx-HD
> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> > Looks very strange for me... 'HD' means half-duplex?
> Yep, I've noticed that too...
>
> > May be linux driver defaults to full-duplex if autoneg fails?..
> Or disabled... And it works -- very strange. And FreeBSD uses
> half-duplex even with given "media-opt" and network is dramatically
> slow -- NFS from DC-local server is about 150KiB/s (from FreeBSD
> installer).
>
I'm not surprised that it doesn't work with autonegotian if autonegotian
is disabled.
If Linux does full-duplex without autonegotiation then _they_ do it wrong
and Hetzner shouldn't rely on wrong behavour.
However since you seem to have access it might be interesting to know the
exact network device and debug why hard setting is broken.

--
B.Walter <> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe. On Wed, Jan 12, 2011 at 12:31:10AM +0300, Lev Serebryakov wrote:
> Hello, Pyun.
> You wrote 11 января 2011 г., 23:00:07:
>
> > rgephy(4) currently always use auto-negotiation to work-around link
> > establishment issues reported in past.
> I think, it is the root of the problem. Autonegotiation is DISABLED on
> these ports. I think, some additional mediaopt (like
> force-half-duplex) for rgephy(4) will be solution.
>
> > For your case, the only way to address the issue at this moment is
> > to use auto-negotiation but that would establish 1000baseT link
> > which would add cost for you. Alternatively request half-duplex
> > configuration to the provider to get a agreed link duplex.
> Maybe, adding new mediaopt is not very hard? Or is it?
>

That had been supported for long time. Just remove full-duplex
media option in your manual configuration.
_______________________________________________
___________________________________________________

Posted on the Freebsd-net mailing list. Go to http://lists.freebsd.org/mailman/listinfo/freebsd-net to subscribe.