ML 310 G5 SATA Raid-1 drive problem

[halftime]

Hi,

I have a Proliant ML310 G5 Server with two SATA drives. The Server has an embedded Raid Controller. The drives are hot swappable and configured in Raid 1 using the Raid utility in BIOS (F8). The Server is running Win2003.

Drive 0 has developed bad spots and I need to replace it.  Can I just hot swap drive 0 with a new drive and the Raid 1 configuration will rebuild it?

I thought I could shut the server down. Take drive 0 out; re-boot the server from drive 1. Plug in a new disk in drive 0 bay and it will rebuild, but this did not work as the server would not boot from drive 1. I had to put the drive 0 back in again and re-boot. I noticed that during the reboot the array has been degraded

The server is still working but I am keen to replace drive 0 as it is reporting bad spots and I know soon it will cause problems. Can someone please tell me how to do this? Any help is much appreciated.

Thanks

[taz_it]

I am using DL360 with hot swappagle drives on Fedora but from what I understand that OS should not matter. What I have done is to just get a replacement drive and (while the machine was running) remove the bad drive and insert in the new drive. The hardware will start the process of rebuilding the mirror automatically (but with all the disk activity the server's performance will suffer). When the light activity settles down you should be back where you want to be with a full mirror.

[halftime]

Hi taz_it

Thanks for the reply.

Drive 0 and Drive 1 are not in sync, if I take Drive 0 out while the machine is running. The rebuilt from Drive 1 will be out of date and not current.

I thought about replacing the Drive 0 after rebuilding the Raid-1 hoping that the 2 drives will sync, then I could replace Drive 0 which has bad sectors but the rebuild process failed.

I may have to replace Drive 0 with a new Drive, recover it from the back-ups. Then plug in Drive 1 and rebuild the Raid-1 again for 2 drives to be optimal. It is a pain as I was hoping that Raid1 will give me redundancy and easy to recover if one disk failed or needed replacing

Just now the Server is working with Drive 0 only which has bad sectors. Drive 1 is out of sync and cannot be rebuilding because the rebuild process fails.

Regards