Replacing failed SCSI HDD in HP Proliant DL360 with Smart Array 5i - RAID 1 cofiguration

[BHR]

I have an HP Proliant DL360 G3, with 2 hot-swap SCSI HDDs in RAID1 configuration.  One of the hard drives recently failed & I bought a replacement to replace it.  I pulled out the failed HDD & put in the replacement, then rebooted the server.  The server booted up sucessfully into the OS with the message:

[787-Slot 0 Drive Array Operating in Interim Recovery Mode
The following SCSI drive(s) should be replaced:
SCSI Port 1: SCSI ID 3

The lights on the new HDD stayed off (while the light on the remaining working HDD remained green).

I booted up the HP SmartStart CD into the ACU according to the HP documentation I found.  I followed the steps until I got to the point where the option to break the (existing) mirror wasn't available as an available task  I attempted variations of this including booting up with the failed HDD installed, the new HDD installed & no 2nd HDD installed (only the working one), but all results were the same.  The logical drive is named "(Interim Recovery) Logical Drive 1", but it doesn't give me the option of "Split Mirrored Array" as stated in the HP documentation, only "Delete" (logical drive), "More Information" & "View Status Alerts".

Restarting the server always only gives the message above, & never any option to recover the failed array.

One odd thing is that when I have the new replacement HDD installed, it is marked as SCSI 1 but the failed HDD is marked as SCSI 3 in the ACU.  The working HDD installed is marked as SCSI 0.

I'm thinking that the reason the drive is not rebuilding is that it is expecting the replacement drive to show up on SCSI 3, but since there is only two HDD bays on the server I don't know how that happened originally or how to address the situation.

Is there any way to manually change the SCSI ID assigned either to the array or to the drive?

[kkpro]

Hi (by the way).

If you problem is really in setting SCSI ID (though I'm not sure), this can be done through jumpers.

http://h20000.www2.hp.com/bizsupport/TechSupport/CoreRedirect.jsp?redirectReason=DocIndexPDF&prodSeriesId=254961&targetPage=http%3A%2F%2Fbizsupport2.austin.hp.com%2Fbc%2Fdocs%2Fsupport%2FSupportManual%2Fc00677580%2Fc00677580.pdf

[gregersenj]

It's a "ghost" Id problem :(.

Problem is, that the old"failed" disk, got a wrong ID. ID 3.

On a DL360 G3, you only got 2 bays - ID 0 and ID 1.

You are most likely running an old FW, and You should get it upgradet. But a FW and BIOS upgrade will not solve your problem.

You have 2 options to solve problem.

1. Delete the old array, and re-install the server from scratsh.

2. Add the new drive as a spare drive. It will forver complaint over the missing disk though.

BR

/jag

[BHR]

kkpro, unfortunately my drive doesn't have any jumpers. It is a model# BD07286224 which isn't on the list. I think the SCSI ID is only assigned by the array controller. I was hoping to find some utility that could do that manually since the ACU doesn't give the option.

gregersenj, I thought that this was most likely the problem. However I don't have any bays left to add the new drive as a spare so unfortunately your #2 option won't work. I was hoping to avoid #1 if there was an easier way.

Thanks anyway.

BHR

[Jimmy Vance]

You might try running the Array Diagnostics utility. I've seen it "fix" odd array settings.

Usual disclaimer - be sure to have a good backup of your data before messing with an array

[BHR]

The ACU doesn't give any options to fix anything.  It doesn't even give me the option th break the mirror.  The only available task that actually does anything it gives is to delete the logical drive.

[BHR]

Would the following work?

• Remove failed drive & install 1st replacement drive into bay 1.
• Remove working drive & install 2nd replacement drive into bay 0.
• (At this point there are 2 new drives installed.)
• Boot up using HP SmartStart CD into ACU.
• Destroy all existing arrays & logical volumes.
• Recreate new RAID 1 array (containing SCSI IDs 0 & 1) & new logical volume
• Power down server.
• Replace 2nd replacement drive drive in bay 0 with original working drive.
• Remove the 1st replacement drive from bay 1.
• Boot the server.
• (At this point would it recognize that bay 1 drive "failed".  )
• Press F2 to fail the missing drive.
• Continue booting up the server (in recovery mode).
• Once booted, insert the 1st replacement drive into bay 1 to rebuild the RAID 1 array.

The thing I am most concerned about is that once I destroy the array + logical drive & recreate it, (a) the server will try to rebuild my good drive essentially destroying all of the data on it; or (b) the server won't boot up with the known good drive since the array+logical drive info on it isn't the same as the array info on the newly rebuilt array.

[Jimmy Vance]

---

@BHR wrote:

The ACU doesn't give any options to fix anything.  It doesn't even give me the option th break the mirror.  The only available task that actually does anything it gives is to delete the logical drive.

---

From within ACU run the diagnostics report

[BHR]

Here (attached) is the ADU report ran from the ACU installed in Windows (if I would have ran it from the SmartStart CD, I wouldn't have been able to save it).

[AndyAY]

Hi BHR,

Have you tried? I got the problem as you and no clue until now!

Andy