MSA 1500CS disk replacement

[Modris Bremze]

Hi.

Have an MSA 1500CS with dual controllers and 2xMSA 20 shelves with SATA disks. One 1TB disk on one of the shelves failed, so a new one was ordered. When the replacement is hot-added in the same slot where the failed drive once was, it also, after a brief green and blue LED flashes, shows an amber LED (failed). The same status also shows up in hpacucli (HP-UX CLI variant). When the disk is put in another slot, no such thing happens - the LED is green, so it looks like the replacement disk itself is ok. Logical drive statuses in the array where the failed disk was, all show "interim recovery mode".

What could be the cause of the amber LED and the replacement not "being accepted"?

[Johan Guldmyr]

Hello, 

bad bay, or something wrong on the connection between the disk bay and the backplane.

That there is something wrong for that port on the backplane is also likely.

Have you assigned the replacement drive to the array/logical drive now?

[Modris Bremze]

Not yet. I figured it would be the next thing to do if the old bay is a no-go. Should I remove the failed drive/bay from the array after I add the replacement drive (which would be placed in a different/working bay)? Would the array rebuild succeed this way?

[Johan Guldmyr]

No, I don't think the smart array is so smart that it knows what you want to do with it.

If you put in a disk in a previously unused bay then I believe it will just stay there until you assign it to an array/logical drive.

[Modris Bremze]

I didn't really expect it to understand that the old bay is broken and I am adding the replacement in another:) I meant to say that I intended to manually add the replacement drive to the same array where the failed drive was and afterwards manually remove the failed drive/bay from the array. Would it be ok with this scenario?

[Johan Guldmyr]

Is the failed drive still in the array? Is the array still doing parity calculations for that broken disk?
Yes, add the drive and sure when it's done take out the broken and put in a blank :)

[Modris Bremze]

Yes, the failed drive/bay is still listed in the array even if there is no actual drive in that bay. The array itself shows up as "Failed" although all logical drives are available to client systems and appear to be fully functional.

I put the replacement drive in another bay and issued:

array A add drives=[box:bay of the replacement]

And got the message:

Error: This operation is not supported with the current configuration. Use the "show" command on devices to show additional details about the configuration.

[Uwe Zessin]

Ah yesss...

The "fun" when running an MSA1500cs with MSA20 attached...

I was aked to help with a case where the "replacement" disk drive did not rebuild in the original slot. I was told that there were already two attempts to replace the disk drive.

Well, the first action was to get the whole thing working again, because somebody did power up the boxes in the wrong order :-(

Then I was asked to bring all software/firmware to a current level...

When I put the replacement disk drive into a second MSA20 attached to the same MSA1500cs, it did show a "prefailure alert". In the end it was decided to replace the MSA20 (still no rebuild) and the disk drive, again.

[gregersenj]

The Smart Array do support drive roaming. But only with a healthy array.

Once you got a degraded array (Drive failure). You must put a healthy drive in the bay of the failed drive.

Pull out the failed drive

Waite at leats 30 seconds.

Insert healthy drive in the bay.

It will begin rebuild within a few seconds (Normally)

I have seen a few times, with old fw and sata drives:

When you insert the new drive, it will lit the amber LED.

After a few minutes it will change to green and start rebuild / copy back.

You can try:

Insert a disk in the bay of the failing disk.

Wait 5 - 10 minutes. Check if it recover.

Upgrade FW.

If it still fail, you have another failure.

backplan

Cable

Array controller

You can add a disk as a hotspare, and it will rebuild to that disk. But it will remain degraded status, even though it hos now regained redundency, untill it get a healthy drive in the bay of the failed disk.

There's a good document called: Smart Array Technology Brief.

BR

/jag