Pro/Forums - View Single Post

jrmvt · 08-15-2005, 01:43 PM

OK, I'm at a complete loss, so I thought I would try running it past the group here.

Back in the Spring, I had my Dell 705N die on me. Finally got it to tell me that one of the four drives in my RAID5 array was bad. Replaced it - still didn't work.

I limped by, storing the data on our SBS2003 server until I had more time to play with it.

Last week was that time. I decided since it was dead anyway and I didn't need the data on it anymore (I had a good backup thankfully), I would do a mind-wipe on it and turn it into a SNAP 4100. I broke the RAID5 set up into 4 individual drives, did the debug trick to make it forget it was a Dell, then proceeded with upgrading the OS. Once all of that was done, I ran the Disk Check on the 4 individual drives and all checked out OK.

So I built the RAID 5 Array and copied the data back over. A few days later, I started getting complaints that a lot of the picture files were corrupt. So I ran the disk check on the RAID 5 Array and got a fatal error while it was checking the cylinder groups. I broke the RAID apart again and ran the disk check on each drive - all registered as OK. Rebuilt the RAID and ran the disk check - Fatal Error!

I checked the Knowledge base on Adaptec's website and it says that for severe errors, sometimes you have to run the Disk check with error correction activated repeatedly. So after 15 successive runs, I was still getting errors. I broke the RAID apart again, thinking that if I rebuilt the RAID 5 array utilizing only 3 of the disks instead of all 4, that maybe I could narrow down if a bad drive was causing the problem. Ran Disk Check on the 3-disk RAID 5 Array and on the remaining disk and got errors on both! I tried again, setting it up so that the one non-RAID disk was the new drive I just added and the 3-drive RAID5 array was the old drives - still errors all around!

So, I'm guessing the problem is one of three things. 1) There are multiple bad hard drives that need to be replaced. 2) The OS is corrupt, in which case I think I'm screwed because the only non-upgrade version I have is a 3.x version from Dell and SNAP does not allow you to roll back from a 4.x version to a 3.x version. 3) There is another hardware error causing the problem (RAM, Promise controller, CPU, etc.)

Anyone have any ideas what I should try next? Thought about pulling each hard drive out one at a time and formatting it on another computer, but I'm not really convinced that the problem won't just perpetuate as I put each drive back in.

Thanks,
J. R.

08-15-2005, 01:43 PM	#697
jrmvt Cooling Neophyte Join Date: Aug 2005 Location: VA Posts: 3	OK, I'm at a complete loss, so I thought I would try running it past the group here. Back in the Spring, I had my Dell 705N die on me. Finally got it to tell me that one of the four drives in my RAID5 array was bad. Replaced it - still didn't work. I limped by, storing the data on our SBS2003 server until I had more time to play with it. Last week was that time. I decided since it was dead anyway and I didn't need the data on it anymore (I had a good backup thankfully), I would do a mind-wipe on it and turn it into a SNAP 4100. I broke the RAID5 set up into 4 individual drives, did the debug trick to make it forget it was a Dell, then proceeded with upgrading the OS. Once all of that was done, I ran the Disk Check on the 4 individual drives and all checked out OK. So I built the RAID 5 Array and copied the data back over. A few days later, I started getting complaints that a lot of the picture files were corrupt. So I ran the disk check on the RAID 5 Array and got a fatal error while it was checking the cylinder groups. I broke the RAID apart again and ran the disk check on each drive - all registered as OK. Rebuilt the RAID and ran the disk check - Fatal Error! I checked the Knowledge base on Adaptec's website and it says that for severe errors, sometimes you have to run the Disk check with error correction activated repeatedly. So after 15 successive runs, I was still getting errors. I broke the RAID apart again, thinking that if I rebuilt the RAID 5 array utilizing only 3 of the disks instead of all 4, that maybe I could narrow down if a bad drive was causing the problem. Ran Disk Check on the 3-disk RAID 5 Array and on the remaining disk and got errors on both! I tried again, setting it up so that the one non-RAID disk was the new drive I just added and the 3-drive RAID5 array was the old drives - still errors all around! So, I'm guessing the problem is one of three things. 1) There are multiple bad hard drives that need to be replaced. 2) The OS is corrupt, in which case I think I'm screwed because the only non-upgrade version I have is a 3.x version from Dell and SNAP does not allow you to roll back from a 4.x version to a 3.x version. 3) There is another hardware error causing the problem (RAM, Promise controller, CPU, etc.) Anyone have any ideas what I should try next? Thought about pulling each hard drive out one at a time and formatting it on another computer, but I'm not really convinced that the problem won't just perpetuate as I put each drive back in. Thanks, J. R.