PowerEdge 2500 died - BSOD followed by 'NTdetect failed'
Posted on 2008-10-28
On Sunday afternoon, I popped into the office to change a few email settings on our Dell PowerEdge 2500 running Windows SBS 2000 and Exchange 2000.
I switched on the monitor to be greeted with the blue screen of death.
I powered it down and started it booting back up, which it did, although one of the RAID 5 drives is flashing the 'X' LED, suggesting that drive has had it.
After booting, the Exchange mail stores were all offline and wouldn't come back online.
After a lot of messing around and a restore of the mail stores, I got them back online with no loss of data. However, the routing tables weren't working and as such, no email could be sent or received. I followed an article on Microsoft's support site which said to reinstall Exchange which would repair the routing tables.
I did this and it asked for a reboot to replace the files that were in use. I hit OK and after running through the BIOS and RAID settings, instead of booting into Windows, it came up with 'NTdetect failed' and nothing else. It just sat there.
I searched the net, which turned up results suggesting the boot files were corrupted and that I should boot using the Win2K CD and use the repair option to replace them. I tried this, but although the LED flashed on the CD-ROM drive when a CD is inserted, the disk isn't actually recognised and it won't boot from it, so I assume the drive has also had it.
I built a bunch of boot floppies and booted into Win2K setup from those. When I chose R to repair, it told me there was no hard drives available with a Win2K installation and quit.
This was at 2am yesterday morning, 2 hours before I had to leave to go to Holland, so I had to leave the office with webmail only and no historical email.
I'm now in Holland and have managed to get net access, so I thought I'd do some digging and see if I could formulate a plan of action before I get back - I'm back on Friday night and need to spend the weekend getting the email back up and running before Monday!
Thinking about it, I think the late hour and stress had made me forget to load any RAID drivers when booting into Windows setup, which is why it couldn't find the installation. So, my first plan is to download the drivers and stick them on a floppy to load in during the boot process. Hopefully that will allow me to do the repair and get back into the server.
I'm hoping it's then just a case of replacing the hard drive and allowing the RAID to work it's rebuilding magic!
The C: drive is a partition on the RAID array, so I assume that the hard drive failing has somehow buggered that? If so, I guess it's a partition error? Is there any way of recovering from that?
Has anybody got any good advice on recovering from this NTdetect error?
I desperately need to recover this server - not just because of the email (which is backed up and, worst case, would only lose a few days worth of email); but because there is a very important database on there - this should also be backed up, but I'm not sure how recent the last back up of it is.