ESX Network redundancy failed, no faults in UCS

[lorenzobexer]

Hi,

we are currently having some issues with UCS. We have 6 UCS servers with ESXi 4.1 installed. All of them seem to work fine, except one ESX is loosing Network redundancy one to three times a day for about one minute. It is always the server located in Slot 3, that keeps failing we already tried putting another server into that slot and associating another ESX server profile to it. Also, the VNICS failing in ESX are always the same: vmnic1, vmnic3, vmnic5 so it seems like there is a problem with the blades VIF path to Fabric B.

This are the VIF paths of the failing Server

And here the VIF paths of Server 1, wich works without problems:

Am i right that there is a problem with FEX Host Port right/3 or do you have an idea how to locate the problem?

Best regards

[cpaggen]

If vmnics 1, 3 and 5 are "routed" to Fabric B and all fail simultanesouly, this indicates an issue with either the network uplink(s) between Fabric B and the network infrastructure, or between the IOM port the server is pinned to and Fabric B. If you lose all network uplinks on Fabric B this brings down all server-side vnics on Fabric B by default. Likewise if you lose the IOM-FI link your server is pinned to this affects the vnics routed to Fabric B.

[lorenzobexer]

Replacing IOM 2 seems tho have resolved the issue.