01-10-2016 02:07 PM - edited 03-01-2019 12:32 PM
I had an environment, where all blades were associate and powered down (customer did some power environment tests). All the OS are installed on local disks (ESXi, Windows, Linux).
Then a power fail happened. Power comes back, and the discovery of the chassis was totally screwed up, the numbering changed.
This implied that powering on the blades resulted in a total chaos; Failure, because there was no blade in this chassis slot; or wrong OS was found.
Luckily, the default scrub policy was Disk = NO.
We had to decommission all the chassis, then recommission them in the old order....
Why is there no persistency on the chassis, so it remembers its number, and comes back up with the original id ?
Solved! Go to Solution.
01-11-2016 04:16 AM
Walter,
That would do the trick. Thanks for replying before I tried to replicate it this morning :). Yeah I noticed on the other posts about the upgrade failure occurring, I wasn't sure if this was related to another environment.
Hopefully everything else recovered and you are back online. If you need any assistance let me know.
Regards,
Qiese Dides
01-10-2016 09:21 PM
Hi Walter,
This is a great concern you raised. Let me do a quick recreate in my lab environment and some research regarding this issue. If I can't find something of a "persistence on the chassis" I will go ahead and create an enhancement request to get this taken care of.
Thanks for bringing this to our attention.
Regards,
Qiese Dides
01-11-2016 02:52 AM
Hi Qiese
I have an explanation:the upgrade 2.2.3d to 2.2.6d failed. see other blogs.
On reboot, the FI ended up in boot prompt, and as a matter of fact, both FI's had to be recovered.
Therefore the configuration of the chassis is gone (same as one would do a erase configuration).
I think any restore (binary or configuration) will not fix the situation.
Therefore in the case of a total failure, one would have to disconnect all chassis.
Then recover FI's.
Restore configuration.
Bring up each chassis one be one, according the proper order.
Thanks guys.
Walter.
01-11-2016 04:16 AM
Walter,
That would do the trick. Thanks for replying before I tried to replicate it this morning :). Yeah I noticed on the other posts about the upgrade failure occurring, I wasn't sure if this was related to another environment.
Hopefully everything else recovered and you are back online. If you need any assistance let me know.
Regards,
Qiese Dides
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide