Problem upgrading VF FC firmware using VCM 2.30 & VCSU

[The Brit]

This thread describes a problem and provides a solution. It doesn't require any responses (unless you feel the need).

The problem was that upgrading VC using VCSU and "vcfwe1f230.bin" failed to complete the upgrade on 4 x HP 4Gb VC-FC Modules installed in bays 5, 6, 7, and 8. The modules were previously at version 1.32, and the upgrade was to bring them to 1.40.

During the upgrade process, VCSU indicated that the firmware package had been uploaded to the FC modules and the activation phase was attempted on all modules. No error messages were generated, all that was displayed was a message indicating that the upgrade had failed on all of the FC modules.

Playing a hunch, and based on the fact that the F/W upload appeared to have succeeded, I tried cycling the power (via software) on one of the modules. This did not appear to have any affect on the F/W version as displayed in the OA screens.
The second approach was to physically remove the module and then replace it. This caused several things to happen.

1. The remove/replace was recorded in the SYSTEM Log.
2. The module was no longer displayed on the "Interconnect Bays" screen, and remained absent for ~3-4 minutes.
3. When the module reappeared, it showed up with the new F/W, i.e. Version 1.4.

After performing this procedure on all of the delinquent modules, they all showed in the OA screens as being at FW version 1.4. In the VCM Firmware Management screen however, only some of them displayed the new firmware (note:, before cycling the modules, none of them were indicating a firmware level as installed/running).

(Note: this clearly implies that a software power cycle IS NOT equivalent to a remove/replace of a module.)

To try and clear this up, I failed over the VCManager, which made all of the modules look the same, i.e. the FW version disappeared from all of the modules in VCM (go figure!).

Anyway, the OA still confirmed that all VC FC modules were running version 1.40, so I quit while I was ahead.

While I am a big fan of Virtual Connect, and HP's blade systems generally, I would be so much happier if they were a bit more robust. I find it very disconcerting when programs produce different outcomes, in virtually identical conditions (no pun intended).

Dave.

[marcus1234]

nice write up thanks
as i replied in an earlier query posted


I am trying to get some one in 3rd level to verify my hunch I have often concluded a reseat is not the same as a reset
even though at times the OA shows the new firmwares,
I have concluded it may be showing them but I am sure in certain configurations they are not actually being used ,

thats what the reseat is preferred I see theses issues on a daily basis.



"i have often seen the similar issues

only seen issues like this when the very old firmwares were updated to recent ones.


worth a try
sounds like a bad flash,,remove both vc modules
...>>>cli reboot OA
then install one only , update that then try other see what happens"

[The Brit]

See above.