Raid 5 setup with 2 spares - how do I .....

[Diane Petrie]

I had a raid 5 setup with 2 spares. When 1 drive failed, 1 of the spares kicked in, leaving me with 1 spare. I have replaced the bad disk, but cannot see how I can specify the replacement drive as a spare so that I can have 2 spares again. Any ideas?

[TTr]

Usually when you replace the failed drive, the "sparing in" process starts automatically and can take several hours depending on the size of the drive and the activity on the storage system.

By now or within the next couple of hours the sparing in will have finished and you 'll be able to see the two spare drives.

If this does not happen, you have to tell us more about your storage system.

[TTr]

I did not explain,

"sparing out" is the process by which a failed drive is replaced by the spare in a RAID group. In this process the spare drive is built from the remaining drive(s) of the RAID group. This is what happened when your drive failed.

"sparing in" is the process by which the spare drive is copied (not rebuilt) back to the replaced drive. This is what is happening right now in your storage system. If you look at the spare and the replaced drives, their activity lights are blinking nonstop.

[Diane Petrie]

Disk 0,1,2,3 were used for data in 1 logical drive, disk 4 and 5 were the spares. Drive 0 failed and disk 5 was incorporated into the logical drive automatically. I replaced the failed drive (drive 0) and I am now showing a status of o.k., and the array shows unused space. I can create another logical drive with it, but I don't want to - I want the space to be used as a spare. I'm attaching 2 screen shots that will hopefully explain it better.

[TTr]

The second screenshot shows all six drives on the left column but in the center "Physical drives" section, drive 0 and drive 5 are showing up as data drives. Are you sure the sparing-in process is not still running? The "percent Rebuild Complete" is not updated.

The spare drives section is cut-off in the screenshot but I assume that disk 4 is the only one showing in there.

I don't have access to an ACU right now and can not provide anymore details.

Two other questions.

How long has it been in hours when drive 0 failed? When drive 5 was incorporated in the logical drive, the rebuilding of drive 5 can take hours and sometimes days. Until drive 5 is completely rebuilt, the replaced drive will not be used.

[Diane Petrie]

O.K. - your response made me go back and look closer to the messages that were generated during the error and recovery. Originally, I was going on the information that was provided to me by my technician. When I look at it, I see that the drive 4 (the spare) was made active and rebuilt after the failure and then made inactive after drive 0 was replaced and rebuilt. All of that seems fine. I now believe (and would appreciate your opinion) that my tech never installed disk 5 properly - that it has been sitting as unused space in the logical array since it was installed last July. Is there any way (without redoing the whole array) that the drive (and unused space) can be removed from the logical drive and added as a spare?

[TTr]

I also took a second look at the drives and sizes and here is the latest.

Drives 0,1,2,3,5 are all data drives in a RAID-5 array group, the array-A. Array-A has 104GB in logical drive 1 and 43GB unused. The used and unused space is spread out on all 5 drives. Usually you can NOT take any drives out of an array without rebuilding it. This would be a function of the array controller.

Also depending on your array controller, what you can do is expand logical disk 1 to use the remaining unused space. But don't do it before you check if your OS can see the increase in the size of the LUN as a drive partition increase.

[Diane Petrie]

Will do. Thanks so much for your help.

[TTr]

One other thing you can do is create logical drive 2 on the unused 43GB space. This would be presented to your server as another disk partition. I don't know if that would be useful to you as a separate partition.

[Diane Petrie]

Thanks. I knew about creating the 2nd logical, but that doesn't help me. I don't really need the space right now, but when I do, I was going to look into using diskpart.exe. I'm running windows server 2000, so I downloaded it from Microsoft so that I'll have it when I need it.