LSI 9271 on C240M3 constantly reporting "Unexpected Sense"

[Steve Galambos]

I have a relatively new (2 months or so) C240M3, and I happened to be going through the CIMC today and noticed the storage log has the following entry from the RAID controller every 3-10min:

"Unexpected sense: Encl PD 08 Path a80c0dc0ce0000bd, CDB: 1a 08 1c 00 ff 00, Sense: 5/00/00"

the CDB alternates between the one above and "1a 00 08 00 ff 00" every so often.

From my interpretation of the message and a little Googling, I think it's telling me there's a new drive in slot 8 (or PD08), however I have no drive in slot 8, the Physical Drive Info tab in CIMC confirms that the slot is empty, and my physical disk present count is correct. I did notice that my Enclosure Device ID is 8 under Drive info and checked all 14 drives for any other info that matches the message, my OEM SSDs have SAS paths of "a80c0dc0ce00008a" and "a80c0dc0ce00008b" which is very close to the "Path" listed, but not a match, and none of my SAS drives are even close to it.

Any ideas on what this message is?

[Keny Perez]

Hello,

What CIMC version are you running?

-Kenny

[Steve Galambos]

Kenny,

The CIMC is running version 1.5(4b).

Thanks,
Steve

[fenilantony]

Any new workaround came for this issue?

[Steve Galambos]

Antony,

I hadn't looked at this in a while, so I just checked everything out this morning. I'm running the newest CIMC version (2.03i) and vSphere 5.5U2 with all but last week's bugfix update installed so unless there's a change in vSphere 6 I don't think there have been any updates to fix it.

I did take a look at the bug, and they have a workaround listed that I haven't tried:

For now, you can try the following on your test system:
1. Open the ESX console
2. Run this cmd: "esxcli storage core device list" and look for the device that has "Enclosure" in its name.
3. Note down it's identifier, usually starting with naa....
4. Run this cmd: "esxcli storage core device set --state=off ?d naa.x" but replace naa.x with the your device identifier.
This should eliminate the messages from the vmkernel.log and Cisco IMC Storage log.

[fenilantony]

I did it,and eliminates the error message.

But i don't think this is a good solution.

[Steve Galambos]

I just upgraded the CIMC firmware to the latest version, 2.0(1a), and I'm still seeing the messages from the RAID controller. I have a maintenance window on Monday so I'll upgrade the BIOS and controller firmware to the latest versions as well, and see if that resolves it.

Thanks,

Steve

[Keny Perez]

Steve,

It is ALWAYS recommended to upgrade the whole system to the same version(talking about C series servers), if you see the issue is not gone after you make all the firmware match, please run the diagnostics tool embedded in the SCU, see below for more info:

http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/sw/ucsscu/user/guide/31/UCS_SCU/booting.html   <<< Downloading and Launching the SCU

http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/sw/ucsscu/user/guide/31/UCS_SCU/diagnostics.html  << Using the Diag tools ...

There is an specific "RAID Adapter Test " so we can rule out the controller.  And there is also a "Comprehensive Tests" (keep in mind this last one can take up to hours) where you can have "Smart Disk Test" that will test the available disks in the system by reading each disk block by block and "VDisk Stress Test" that will stress the virtual disks in the system.

I hope this helps...

-Rate ALL helpful answers

Kenny

[fenilantony]

My C240M3 also generating an error "Unexpected sense: Encl PD 18 Path 4c4e3544245e00bd, CDB: 1a 00 08 00 ff 00, Sense: 5/00/00"

Do u have any idea about this error?

Cisco Integrated Management Controller (CIMC) Firmware Version:1.5(4e)

[Keny Perez · Accepted solution]

I think you might be hitting this bug  (as long as you are running the recommended drivers for the controller):  https://tools.cisco.com/bugsearch/bug/CSCuj10535   ; It reads "VMware seems to treat all storage devices the same way, regardless of whether they are SAS disks or just enclosures. The messages you are seeing means that the host (ESX) was sending mode sense commands to the enclosure and the enclosure does not give a valid response because mode sense commands are not a command that should be sent to an enclosure. So the unexpected sense messages from the enclosure are benign info messages."

Check the workaround:

"

There is way to "disconnect" the enclosure from the viewpoint of ESX. It should not have an adverse impact as the enclosure is not a usable storage device.

For now, you can try the following on your test system:
1. Open the ESX console
2. Run this cmd: "esxcli storage core device list" and look for the device that has "Enclosure" in its name.
3. Note down it's identifier, usually starting with naa....
4. Run this cmd: "esxcli storage core device set --state=off ?d naa.x" but replace naa.x with the your device identifier.
This should eliminate the messages from the vmkernel.log and Cisco IMC Storage log."

HTH

-Rate all helpful answers

Kenny

[Keny Perez]

BTW, these messages should not mean any problem to the controller...

-Kenny