HP MSA2000 Enclosure Replacement

[Speculator]

Hi,

We're running a HP MSA2012i supporting our VMWare infrastructure.  The device has a second disk enclosure, which has recently experienced some problems.  HP support have decided to replace both PSU's in the enclosure and the midplane itself.  This is essentially going to mean replacing everything apart from the disks as far as I can make out.

I wasn't able to get a convincing answer from HP support, so I was hoping someone here might know the answer to these questions:

Firstly, is changing the midplane in this enclosure going to effect my existing vdisks at all?  That is, will I just be able to plug it back in and have it recognise the disks, the RAID setup, vdisks and volumes on the replaced enclosure just as if nothing had changed?

Secondly, do I need to power down the main unit to change these enclosure parts, or can I hot-swap this disk enclosure for another one (providing all the VM's running on it are powered down)?

Thanks in advance! :)

[hvhari · Accepted solution]

meta data for vdisks & configuration is stored on the controllers & disks themselves. Replacing ONLY midplane will not have any effect on these.

Instructions of Midplance replacement is available at : http://bizsupport1.austin.hp.com/bc/docs/support/SupportManual/c01657200/c01657200.pdf

[Torsten.]

IF you change the expansion chassis midplane, nothing will change. This is different if you change the main enclosure!

Do it offline, otherwise you will loose parts of your disks during the exchange (= data loss. maybe).

[Speculator]

Hi,

Thanks for both replies.  Just to be clear, when you say "do it offline", you mean to power down the entire SAN, right?

L8r.

[Torsten.]

Not the entire SAN, but the MSA2000.

[Speculator]

Sorry, I wasn't being clear.

I meant the MSA, as it's virtually our entire SAN anyway.

As it turns out, I didn't have that luxury.  We'd scheduled the downtime to do the work this evening, but the enclosure failed before we got the chance.  Both PSU's failed within a minute or so of each other taking down the whole enclosure.  I powered it back on, and then within minutes it had done it again.  So, figuring it was down anyway, I swapped out the enclosure and PSU's for the new parts HP had sent.

So far, it's running okay with no problems (it's only been half an hour though).

Just in case anyone is interested, the problem was the PSU's shutting themselves down with an error "power supply status  Under voltage, DC failure".  Which was incorrect.  Removing power from the PSU and reapplying it caused the unit to return to functioning normally again for a varying length of time (ranging from days to minutes).  It started on one PSU and seemed to spead to the other until both were failing every few minutes.

Anyway, thanks for the input from everyone.  It's much appreciated.  I should also thank Microsoft and VMWare somewhere public too, their filesystems have proven quite resilient to SAN storage that kept failing.

L8r.