RAID 10 failed drive will not automatically rebuiild itself HP Proliant ML370 Smart Array 431

Hello, have a little Hardware RAID issue. I have a HP Proliant ML370 G2 server that has a hardware raid controller, Smart Array 431. On this controller I have two arrays (A and B). Array A is a 3 drive RAID 5 configuration using SCSI ID 0-2. Array B is where the problem lies. Array B has a 2 drive RAID 10 configuration. One of the drives has failed and when I went to put my spare drive in the array would not automatically rebuild. I noticed the server never had any of the HP utilities on it so I downloaded the Array Configuration Utility from HP thinking it would have an option to rebuid the array but it does not. At first I thought I had a bad spare drive so I replaced it with a second drive which does the same think. When I look in the ACU utility it looks a little funny for me. The Array B says it has a disk on SCSI ID 3 which is the workiing disk and a failed disk on SCSI ID 6 but there is not a SCSI ID 6 on this server. The drive is in the next physical slot on the drive bank so I would think it would be SCSI ID 4. The new drive I put in shows in the ACU utility as being on SCSI ID 4. Has anyone ever seen this and how can I force the array to rebuild?
jffisherAsked:
Who is Participating?
 
lnkevinConnect With a Mentor Commented:
The mirror set of drives only contains data as teh OS is on....

If it is the case, full backup or copy your data out of that raid container, delete the current RAID 10 and recreate a new raid container, recreate the partition and copy your data back. It's much easy without OS or system file involved.

K
0
 
lnkevinCommented:
Backup your data to elsewhere before perform the task to mimize risk. Get in ACU, on the lower right, click on Configuration Wizard. Follow the wizard and make sure you chose the RAID 1 container, click on the check box next to the new drive to select it in RAID 1 container. Finish with the wizard. Your array will not be changed until you confirm in the end of the wizard. You can always go back and undo the change before exit ACU. Be confident in doing this. ACU will not delete your data without giving you some warning.

K
0
 
jffisherAuthor Commented:
Inkevin
Is this a rebuild feature or am I creating a new RAID10 configuration.
Also, will this erase the data on the disk? The mirror set of drives only contains data as teh OS is on the RAID 5 disks. I do have the data backed up using Veritas backup exec doing a disk to disk backup with weekly fulls and daily incrementals.
0
Improve Your Query Performance Tuning

In this FREE six-day email course, you'll learn from Janis Griffin, Database Performance Evangelist. She'll teach 12 steps that you can use to optimize your queries as much as possible and see measurable results in your work. Get started today!

 
andyalderCommented:
What must have happened is that previously a SCSI ID line failed on the disk (or due to a dirty backplane connector) and the Raid Information Sector on the disks has been automatically updated to reflect this. Now the rest of the disk has failed and the controller wants a disk to be put at ID6 to match the relocated disk. It's not the first time this has happened, I've seen it posted at EE a few times in the past. Unfortunately there is no easy way out of the problem - you can't add a hot spare after failure and you can't update the RIS to tell it to use the disk in bay 4 instead.

I can even reproduce this fault on my test server on which I have cut a track on the backplane.

There is one way that migt work, with the latest ACU there is an option to break the mirror which leaves you with a RAID 0 array of one disk. You may then be able to add another disk to the array and migrate it back to RAID 10.
0
 
jffisherAuthor Commented:
Inkevin,
that is what I was thinking and maybe even doing a robocopy first to an external HD just in case.

andyalder,
I downloaded the latest version of ACU and did not see the optoin to break the mirror, where would that be located (when you enter the wizard). I was hesitant on entering the wizard in case some data may be lost. Is it correct that no changes are made until I choose to save the configuration.
0
 
lnkevinCommented:
Is it correct that no changes are made until I choose to save the configuration...

Yes, it is. One advantage of using ACU like I said, it wont delete your data without giving you some servere warning!!!

Robo copy or Xcopy will work in your case. Just copy the data out of there first and you have more confident to mess with ACU. You should delete the container anyway.

K
0
 
andyalderCommented:
Proceedure is at http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00378986/c00378986.pdf, you may need to upgrade controller firmware. Backup/restore might be quicker.
0
 
jffisherAuthor Commented:
andyalder,
Looks like my smart array 421 only supports the manual mirror breaking method but the bigger problem is that I have 2 arrays on the one controller and using the manual method requires me to break all array configurations in order to fix my mirror. I think on Saturday, with my data backed up, that I will add another drive and re-create the array doing a raid5 with 3 disks and 1 logical drive unless someone else knows how to force an array to rebuild intself.
0
 
jffisherAuthor Commented:
Correction, smart array 431.
0
 
lnkevinCommented:
Unfortunately, I don't have a failed drive to test your case in my lab. I can only suggest you to backup delete and rebuild the container.

K
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.