Monitor physical NIC state?

[robincm]

Hello,
I've recently had a VMware ESXi 5.0 host that started experiencing intermittent network connectivity on the uplink from a vDistributed Switch. The cause was a dodgy patch panel.
I noticed it because some VMs were randomly not pinging via our other monitoring solution.
vCenter has got the link state events showing in the host's events, but I've not seen anything via Veeam One - should it have picked this up by default?

Is there any way that I can get this monitored?

Events look like this in the vSphere client:
Physical NIC vmnic8 linkstate is up. info 11/09/2014 14:35:58
Physical NIC vmnic8 linkstate is down. warning 11/09/2014 14:35:46

Thanks!

[Vitaliy S.]

Hi Robin,

Yes, you can monitor these events with Veeam ONE as well. Just add either event names or their descriptions when creating new alarm via alarms creation wizard in Alarms Management tab, and then assign newly created alarms on the vCenter Server/ESXi host.

Thanks!

[robincm]

Hi, what is the event name? I did do a searching for this but the event name seems to be quite hard to determine?
Might it be this: esx.problem.net.vmnic.linkstate.down? (found here https://communities.vmware.com/thread/469457)
The Events display in vSphere doesn't have a "Name" column, only a "Description" column.
Is there anywhere I can get a list of these event names from?

[Vitaliy S.]

Veeam ONE by default retrieves all events from the vCenter Server.

Might it be this: esx.problem.net.vmnic.linkstate.down?

Yes, it looks like this is the one you need, but to confirm that can you open Veeam ONE Monitor Client -> Events tab and then find the event that was generated during physical NIC failure? Once you do this, you can check out the name of the event and use it as an alarm trigger.

[robincm]

Thanks.
Yes, events are esx.problem.net.vmnic.linkstate.down and esx.clear.net.vmnic.linkstate.up
I've created one rule that generates an error on the "down" event and "resolve" on the "up" event. I'm hoping that will therefore allow these to self-resolve. Sound correct?
Perhaps this should be a Predefined alarm? It's not good if you loose connectivity on a network uplink.
Thanks very much for your help and speedy response (as usual)

[Vitaliy S.]

Yes, that's correct and yes, this should be a predefined alarm too. Thanks!

[robincm]

Do you want to add it to the vNext wishlist or shall I?
Having tested it this morning, that all works fine, but I've actually decided to remove the self-resolve on this alarm - if I have any loss of connectivity to a vmnic I want to make sure it gets flagged and stays flagged until somebody has determined what the cause is.

[Vitaliy S.]

I will ask our R&D team to do that, no worries. As to the removing the event that resolves it, I think you still can keep it, but use email notifications to the Admin group with a date, time and host information of the failure for troubleshooting purposes.