Problem with a lan interface card

[Trenta Angelo]

Hi all,
I have two clusters (both of two servers L 1000), 1) prodcc03 and prodcc04
2) prodcc05 and prodcc06.
Listing of command lanscan was the same, with lan3 and lan4 down, but now on p5 lan3 is up, and I didn't remember I had changed it.
Are there problems?What I can do?

Regards, Angelo.

[Sanjay_6 · Accepted solution]

Hi,

Seems like one of your network card /cable might have failed on the server where you have this problem.

Do a "cmviewcl -v" to look at the cluster information.

use
cmviewcl -v -n node_name to look for cluster information for only one node (node_name).

Look into /var/adm/syslog/syslog.log file to check if cmcld has logged any information about lan switching.

Hope this helps.

Regds

[Sridhar Bhaskarla]

What does your ifconfig lan3 say?.

Did you check your syslog.log to see if there was a local lan failover to lan3 and then failback as lan3 being your standby?.

-Sri

[Christopher McCray_1]

If I understand correctly...

No, there is nothing to worry about. Your lan3 and lan4 are your standby lan cards in your cluster configuration. They are used in the instance where you experience card failure on one of your primary lan cards, where MCSG will then configure the standby with the ip/subnet mask of the failed card and continue on. There could have been any number of factors as to why they were down before. Was there any network configuration being don on your lan recently, or are these new cards? Did you have to get replacement cables installed? Just some thoughts, but overall, you needn't worry.

Hope this helps.

Chris

[Trenta Angelo]

I find this informations inside syslog, and other commands you suggested me.
May you check it?

Thanks, Angelo

[Christopher McCray_1]

Apparently something occurred with your lan card causing a hardware reset, which was detected by the cluster. It is complaining about auto-negotiation, which since it's a 100base-t, the port on the switch is not nailed down to 100Mb. I have only heard about this in where you will want to hard-set the speed to what your using. In any case, this happened to one of your standby lan cards, so you undoubtedly experienced no down time. I would have a talk with whomever manages your network to see if anything happened on his/her end around that time. Also, check the configuration of the lan card because when the card was reset, It did so to half-duplex, you will want to change it to full.

Hope this helps.

Chris

[Sanjay_6]

Hi,

It is okay if you see the lanscan status as up for all the cards configured in the cluster. In a cluster all the configured ports are in up state and the cluster will switch to secondary ports in case one of the primart port fails.

So for all the ports you see in cmviewcl for a particular node, it should show the status as up unless the port/cable has failed. The lanscan output for the same ports should also show as "up".

Hope this helps.

Regds