"Unknown" status of an empty Blade bay (RESOLVED)

[chuckk281]

Kelly was looking for some suggestions:

**************

Looking for ideas…..

I have an enclosure which is very difficult to get a FULL enclosure outage.  I can lighten the load, and move a few items, but to get the full enclosure offline – if challenging.  We have also had a few unexpected enclosure level disruptions recently, so I am very concerned about what steps we try to clear our problem.

I have an enclosure where a VM host started to act up (loosing connectivity), the customer placed it in maintenance mode Saturday night and tried to reboot it.  It then went to a “Failed Profile” (Red X), we unassigned the profile (profile went Green), then re-assigned the profile to the blade (back to a Red X).  I then asked them to do an E-Fuse on the blade, the OA reported – “E-Fuse cannot be reset”

Since we could not E-Fuse it, I suggested to re-seat the problem blade (bay 15).  After a re-seat, it still would not power up.  We then moved the blade to another enclosure where it powered up just fine (BL460 Gen8).

With the blade removed from the original enclosure/bay 15 and the profile unassigned, the enclosure is reporting “Unknown” status for that bay, which is empty.  So – I believe my problem is with the OA communication/enclosure.

I suspect my next course of action will need to be reset of the OAs, and possibly followed by a re-seat the OAs and OA tray.  I know this should not be disruptive, but I have had very poor luck recently. 

This enclosure is currently running 3.56 OA code and the VC is at 3.51.  We are planning to upgrade the firmware this week, but I am concerned with this status if there are some other problems going on and could hit other issues.

If anyone has some insight or suggestions, I am open to hear them.

***************

Input from Mike:

Try to reset OA2 (passive) and perform a failover from the primary OA1 to OA2. Also check enclosure firmware and make sure they comply to SPP. Verify OA2 and VC functionality before reset OA1. Also check if OA2 can manage the bay 15 blade.

And perform failback to OA1 if required. 

You may want to consider to backup the config of OA and VC before the reset.

Radi suggested:

It make sense to check the show all from the enclosure too. Look for Internal data  errors…

Reply back from Kelly:

Just an FYI – we fixed our “Unknown” status of the empty bay.

We reset the non-Active OA first, once it was reset, we made it the Active OA, then reset the other OA.  After that - the bay showed empty.

With the bay showing empty again via the OA interface, we re-inserted the BL460 Gen8, and everything is working fine.  OS is back up and running, no networking issues.

***************

Comments or questions?