Cache Degraded for UCS-C240-M3S Server

[mohamedbineesh]

UCS Server showing Cache Degraded Error. 

[Kirk J]

Greetings.

Would probably need to review server techsupport logs to see what is triggering that.

Drive failures, and subsequent VD rebuilding/healing will also put the cache into writethru mode.

Occasionally you can have a BBU/supercap failure, that the GUI is not reflecting.

Maybe run the 'learning' cycle another time, and see if it clears.

Kirk...

[mohamedbineesh]

@Kirk J   Hi kirk, Thanks for your response,

 I have made a "start learning cycle" under BBU and it completed successfully.(screenshot). However, the Cache degraded error is still there.

So where could be the issue?  The issue could be the Transportable Memory Module (TMM) itself and not BBU?   As per my understanding, the TMM itself is the Cache Memory? 

From the doc : https://www.cisco.com/c/dam/en/us/products/collateral/servers-unified-computing/ucs-c-series-rack-servers/c240m3-sff-specsheet.pdf

" The 9271CV-8i RAID controller card includes a 1 GB Transportable Memory Module (TMM) and a chassis-mounted (remote) supercapacitor (supercapacitor not available on the 9271-8i) for data cache power backup"

[mohamedbineesh]

Hi @Kirk J,  Thanks for your reply. 

Ran the 'learning' cycle again and battery relearn completed successfully, but the Cache Degraded Error is still there. 

So where could be the issue? Could be the Transportable Memory Module (TMM) itself and not the BBU/Supercap?  I hope the Transportable Memory Module (TMM) itself is the Cache Memory?

From the  Spec Sheet, the statement   "The 9271CV-8i RAID controller card includes a 1 GB Transportable Memory Module (TMM) and a chassis-mounted (remote) supercapacitor (supercapacitor not available on the 9271-8i) for data cache power backup "

[Kirk J]

You are correct (TMM=cache).

You will need to open a TAC case, to have the CIMC tech support evaluated , which will contain raid controller and BBU/supercap related logs.

The SuperCap based cache modules tend to have better life spans than the traditional BBUs, but eventually need to be replaced.

Kirk...

[mohamedbineesh]

@Kirk J 

From the  Spec Sheet, "chassis-mounted (remote) supercapacitor for data cache power backup "   Does that mean Supercap is physically Separate from the RaidController(TMM)? If we have to replace the Supercap we can replace it alone .. and if we have to replace the TMM, then we have to replace the whole Raid Controller Card as the TMM is residing directly on Raid Controller. ?

[Kirk J]

You are correct, on both accounts.

While the TMM can be removed from the raid card, I don't know that I've seen it actually sold separately.

Thanks,

Kirk...

[mohamedbineesh]

@Kirk J   :  Today morning the Status Changed back to Normal. Same thing happened last week also. WB changes to WT and sometimes returns back to WB.

BBU changing from WT to WB back and forth

[Kirk J]

It will do that normally when the learning cycle is running.

But the learning cycle should only run once a month, so there must be something about the SuperCap that is flapping.

It will take reviewing a tech support from the CIMC, and specifically the raid controller tty logs that get pulled that will shed light on what is really happening.

Kirk...