I have an Intel motherboard (S5000PSL) that has onboard RAID 1 configured using the Embedded RAID II Technology Tool. The two drives are Seagate Barracuda 500GB 3 GB/s.
The RAID 1 has degraded, and as a result the server (Windows 2003 Standard Edition) has started to blue screen. We have shut down the machine, replaced the faulty HDD, gone into the RAID Tool during the POST, rebuilt the RAID, but upon reboot, the server has requested a CHKDSK which is failing with an unknown error, and then reboots. We can't get past this stage and are wondering why the mirror has failed to recover.
Have we missed something obvious here during the recovery process? I'm expecting the new drive (master) to rebuild from the copy of the OS and data on the slave, but this doesn't appear to be happening. I've tried contacting Intel via their online chat but there is never anyone available, and local resources are also unavailable.
Also, I'm not aware of RAID being able to run off the slave, or to make the slave the master, but is there a chance that we can use this to get the OS running, and then use the new HDD (same spec's) to become the slave while the rebuild take place?
I'm hoping there is an answer to these questions, and appreciate anyones assistance in getting this sorted out.
Thanks in advance.
With RAID 1, you should be able to boot from either one of the hard drives. Degraded RAID 1 won't cause OS boot failure. However, in you case seems you have corruption in file system.
I suggest you connect the two drives one at a time and see if one of them can boot successfully. If yes you can add a new drive and rebuild the RAID. If you you'll need to fix your filesystem first by booting to OS CD and load F6 driver.
Thanks for your help on this. We put the secondary HDD into the primary slot and rebooted. The OS came back up, but with a couple of TCP errors - DHCP, IIS, WINS services (even though we aren't using it) etc were all failing, but we have resotred the IIS Metabase, and reconfigured DHCP, so it looks like we're back on track. Hopefully we're at the end of our troubles here - bullet dogged, but would have preferred it everything worked first time.
Thanks for the advise as well, we will keep these in mind going forward and use your tips for troubleshooting the next time.