Using off-the-shelf replacement for a failed drive (P4300)

[Jim Ruzauskas]

I have a P4300 that just ran out of contract a couple months ago(unfortuately I missed the renewal), naturally we had a couple drives fail on our RAID-6 array.  I ordered replacement HP drives direct from HP Partsurfer per the spare number the support folks gave me.

The drives arrived today, so I popped them in one by one.  No dice - still show as "Off or Removed".  In watching them as I insert them - the green light comes on briefly a couple times.  Then it goes solid orange.

Does anyone know if there is a special "signature" or partioning that HP applies to the P4000 drives before they ship them out?  This is starting to get frustrating, waiting on a call back from support at this point but I'd love to get that array rebuilding sooner rather than later.

[Jim Ruzauskas]

Nothing conclusive on the off-the-shelf issue, I'm leaning towards there being another issue at play on my node.

From all that I'm reading it should just start rebuilding - unless there is some error.  Checked the ADU log and the remaining good drives have no hard read errors.  Even the P400 Smart Array docs say that you can replace both failed drives simultaneously in a RAID-6 array failure situation.

Talked to support and they said I need to go into "Repair Storage System" and re-apply the RAID config.  Further reading shows this will kill any Network RAID-0 data I have, so I'm getting my good backups in place for that data.  I will first try to re-initialize the controller as I read elsewhere, since that will take much less time than rebuilding / restriping the data if it works.

[Jim Ruzauskas · Accepted solution]

Shutdown the the node having trouble, then powered it back on.  At boot on the console I had a message from the RAID controller about the array status and a prompt asking if I wanted to rebuild and save the data or rebuild from scratch.  So this kicked off the rebuild process.

Frankly I was a little surprised to get that question, you'd think wiping out your data would require more steps (for safety sake).

[Bryan McMullan]

Just for future reference, I've never had an issue replacing the drives with off the shelf HP drives.  There is nothing special about them, though I've never tried using drives that weren't HP (I'm curious if non-hp drives would work).

I believe what you saw is unique to a multiple drive failure situation.  One drive fails, you can just pop and swap.  Two drives fail at the same time and the system needs to you confirm the rebuild.  Go figure....

[taylorb]

A couple years back there was a firmware issue that caused drives to report false-positive errors and fail the drives.  I had 3-4 go bad within a month and then after the firware updates, nothing since.   

So I guess I am saying, check that your firmware is all up to date and read all the advisories for failed drives.