Servers on C210 M2 all become unreachable

[Jeremy McCarthy]

We have 2 UCS C210 M2 servers at 2 different locations. Over the past 2 months, they have each had the servers that are configured on them stop responding (including the ESXi host) and we are forced to reboot the CIMC in order for the servers to come back online. Originally TAC told us it was due to this bug: https://tools.cisco.com/bugsearch/bug/CSCtw68712

We installed the firmware update to each system and they were quiet for a month. Last Friday one of the servers showed the same behavior and we were forced to restart again. TAC is now telling us that the problem is due to a combination of the RAID controller battery needing to be replaced, the ESXi firmware needing to be updated (Version 00.00.04.14.18-1vmw, Build: 348481, needs to be 00.00.05.33) and that the NIC driver needs to be updated (version: 2.0.7d-3vmw, needs to be 2.2.1i.v41.1.

Has anyone run into a similar issue? Did you notice anything else that helped fix your issue?

[Manuel Velasco]

What firmware did you upgrade the server too? 

Did you upgrade all of the server components to this version?

[Jeremy McCarthy]

The firmware was upgraded on both chassis' to 1.4(3c) as shown in the bug fix. We did not feel that upgrading it any further would gain us anything as we did not want to accidentally introduce any new issues to the system.

[vijendra Saravanamuthu · Accepted solution]

though it says it was fixed in 1.43(c) , i remember having the same issue on that release as well.

the key is to disable the interrupt remapping in the BOIS first - reboot

then disable it also in the esxi - reboot.

have you done that?

also i suggest to upgrade to 1.43u or v.  1.4(3c) is not even in the list of firmware to download in teh cisco website.  either its been differed or cisco dont want us to use it.

Thanks

vijay

[kedevara]

Hello Jeremy,

Check this link http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/c/ts/guide/TS_C-Series/TS_2Server.html

Hope this help .