link down up

[Mderin]

Hi All,

The Tru64 server(GS1280) running on 5.1B PK4 has a problem about network link which is down&up randomly.

Comments are appreciated.

Here is the a part of /var/adm/messages file.

Oct 12 14:30:09 efes vmunix: bcm3: Transmit timeout
Oct 12 14:30:09 efes vmunix: bcm3: Link down
Oct 12 14:30:09 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:30:13 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:30:13 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:35:47 efes vmunix: bcm5: Transmit timeout
Oct 12 14:35:47 efes vmunix: bcm5: Link down
Oct 12 14:35:49 efes vmunix: lag: lag0 port bcm5 failed
Oct 12 14:35:51 efes vmunix: bcm5: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:35:51 efes vmunix: lag: lag0 port bcm5 operational
Oct 12 14:36:37 efes vmunix: bcm3: Transmit timeout
Oct 12 14:36:37 efes vmunix: bcm3: Link down
Oct 12 14:36:37 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:36:42 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:36:42 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:36:44 efes vmunix: bcm3: Transmit timeout
Oct 12 14:36:44 efes vmunix: bcm3: Link down
Oct 12 14:36:44 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:36:52 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:36:52 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:36:55 efes vmunix: bcm3: Transmit timeout
Oct 12 14:36:55 efes vmunix: bcm3: Link down
Oct 12 14:36:55 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:36:58 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:36:58 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:37:03 efes vmunix: bcm3: Transmit timeout
Oct 12 14:37:03 efes vmunix: bcm3: Link down
Oct 12 14:37:03 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:37:06 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:37:06 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:37:09 efes vmunix: bcm3: Transmit timeout
Oct 12 14:37:09 efes vmunix: bcm3: Link down
Oct 12 14:37:09 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:37:13 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:37:13 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:37:24 efes vmunix: bcm3: Transmit timeout
Oct 12 14:37:24 efes vmunix: bcm3: Link down
Oct 12 14:37:25 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:37:28 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps, full duplex)
Oct 12 14:37:28 efes vmunix: lag: lag0 port bcm3 operational
Oct 12 14:37:35 efes vmunix: bcm3: Transmit timeout
Oct 12 14:37:35 efes vmunix: bcm3: Link down
Oct 12 14:37:38 efes vmunix: lag: lag0 port bcm3 failed
Oct 12 14:37:38 efes vmunix: bcm3: Link up via auto-negotiation (1000 Mbps

[Ivan Ferreira]

Disable the autonegotiation and configure the adapter manually, you can do it using lan_config -a0 -x1 -s1000 . Add the line to /etc/inet.local for making the changes at boot. Also, set the same speed at the switch and disable auto-negotiation.

[Al Licause]

Disabling autonegotiation is not a very good solution in this case since the bcm driver ignores attempts to disable autonegotiation when operating at 1000Mbps.

Which mode are you using in your lag configuration ?

How many actual nics are assigned to this lag group ?

If more than 2, is it only bmc3 that is experiencing the problem ?

Look at netstat -i and/or netstat -is -Ibcm3 to see if this device is logging errors and of what type.

[Mderin]

After we replaced the card DEGXA-TB with DEGXA-SB, it appears that problem is solved. Maybe , there is an incompability between cisco switch and the old card.

[Ranjith_5]

Hi,

I suggest you try replacing the network cable.

Regards,
Syam

[Mderin]

we did it :)

Thanks .