RAID5 Ready for rebuild?

[Ben Russell]

Dear All,

I seem to be encountering a similar issue but don't usually use HP kit. I've recently replaced a faulty HD on a DL380 and the array won't rebuild. It stays in the status waiting for rebuild.

Please find attached ADU report for the server in question.

I would appreciate any assistence in answering the following:

1. One of the drives is encountering hard read failures which in turn is preventing the rebuild? TRUE / FALSE

2. The only way to fix this is to backup the data and rebuild the array from scratch? TRUE / FALSE

Regards

Ben

[Brian_Murdoch]

Ben,

I can see from the analysed report (See attached file) that you have replaced SCSI ID 0 and that ID2 has lots of errors. This drive (SCSI ID 2) is preventing the rebuild.

You may have to rebuild the whole array from scratch, replacing drive 2 as well as drive 0 but you could try upgrading the Smart Array 6i firmware from 2.48 (very old now) and change the rebuild priority to high in ACU before trying again. Before doing anything however make sure you get a full backup of the system as it's in a bad way.

Regards,

Brian

[Ben Russell]

I've recently upgraded the firmware on the RAID controller and changed the rebuild priority to high but am still not getting anywhere.

I am going to carryout the following in order to get the server back to a normal state and would appreciate any advice prior to carrying out these steps. I've attached the latest ADU report for review.

1. Clone the existing data from the server using Acronis or Ghost Solution suite.
2. Remove the drive with the hard read errors in SLOT 2.
3. Remove the exisiting array and data from the drives.
4. Re-create the array.
5. Clone the data back onto the drives.

Regards

Ben

[j1500]

I have an issue where my drive starts rebuilding when I restart my server. During BIOS it prompts to press F1 to start the rebuild process or F2 to continue and assume potential data loss.

This has happened 3 times over the last month.

Is someone able to look at my Array Diagnostic Reports and see if they have any errors showing up that would lead to a replacement of a SCSI board, drive, or other component?

I really appreciate it!

I included a report which shows during rebuilding and after rebuilding.

[j1500]

Here is the report after rebuilding.

[MT19]

Everytime I've had this issue we've always replaced the SCSI backplane, controller, sometimes systemboard, SCSI cables, updated firmware, updated Proliant Support Pack. But always in the end, none of the above worked, and we would end up rebuilding the array and doing a restore from a backup. No matter what, if something is corrupted in the array, the new drive will not be able to rebuild off it.
You can do the above so you can go back and tell the managers/server owners "I ruled out the hardware and firmware, so now it looks like we need to rebuild the array and do a restore from a backup." OR you can save some downtime and just proceed directly to rebuilding the array and conducting a restore, which is what I recommend.
MT

[KarloChacon]

yes MT19 is right that issue happens since there is another HDD that is causing the rebuild process to fail so RAID needs to be rebuild and another HDD must be replaced

[j1500]

It looks like the drive labeled ID 1 is reporting errors. If I replace this drive do you think the array will rebuild correctly?

[KarloChacon]

OK I took some time to check ADU report

my recommendation backup your data and replace HDD SCSI Port 2, Drive ID 1
after that recreate your RAID


since the other 2 HDDs have Fl Rd Recv (Failure Read recovery Errors) which are causing the rebuild to stop or fail


details:

SCSI Port 2, Drive ID 0
Fl Rd Recv 1

SCSI Port 2, Drive ID 1
Hrd Read 2602608
Rtry Write 234887682
Fl Wt Recv 58833
Pred Fails 234891266
Hrdw Errs 58833
Med Flrs 66
Format Err 43

SCSI Port 2, Drive ID 2
Fl Rd Recv 1


regards

[neutronstar13]

Raid 5 won't rebuild
I have the same issue on a DL580-G2 with a 4 disk Raid5 array. 2 disks started flashing Imminent Failure simultaneously. Tried replacing one but keep getting read errors from the other failing disk and the rebuild stops after about 4 or 5 minutes. ADU says "Ready for rebuild". Tried upgrading FirmWare, no help there.

Told the client that the array would need to be rebuilt, OS reinstalled, and data restored from backup, but we are looking for a quicker solution if one exists.

That doesn't seem likely based on what I have seen in my research, but if you have another answer I would love to hear it.