So awhile back (probably around 3 months ago), my first RAID5 file server (4x160GB) had a hiccup, where it wouldn’t detect 1 of the 4 drives and the Promise RAID controller would enter this “critical state”. I powered off the computer and checked to make sure all the cables were fine and made sure they were snug and tight. I rebooted, it started working again. Of course it needed to rebuild the drive (which takes ~2 hours) and after the rebuild was successful, I thought nothing much about it.
A few weeks ago, I noticed it was once again in critical state. To clarify, I hardly log onto this machine as it only serves as a file server, so I have no idea how long it’s been in critical state, but I can only assume it’s been for quite some time. I rebooted the machine several times and it finally detected the drive and started to rebuild. This time, after rebuilding, using Promise’s PAM (Promise Array Management) software, I told it to synchronize to make sure everything was working fine. When I came back to check on it, I noticed it kicked that drive off the array again.
I decided to move all the data onto my second file server (yes, I know that I promised to blog about that and yes, it’s still coming. ETA: undetermined) as a prevention measure. The 2nd file server is a 2.5TB and had over 50% free so it can easily accommodated a 480GB file server, which was nowhere near full.
I finally got around to testing that file server. First I tried changing SATA cables, but it pretty much failed immediately after rebuilding. I wanted to avoid running diagnostics on the drive since this was a hardware raid controller, the individual drives weren’t expose to the controller, which meant it required me to take the drive out and plug it into a different machine (since this motherboard doesn’t have on-board SATA ports). I finally caved and took the drive out and plugged it into my main box.
I installed Seagate’s SeaTools for Windows and pretty much 5 seconds into the short test, results show FAILED. Sigh… Looks like I’m going to need to RMA the drive. Interestingly enough, I was still able to format the drive. This is my first RMA in years. The last one was probably back in college?
I’m not saying Seagates are bad. In fact, 90% of the hard drives in my computers are Seagates and the other 3 drives on my Promise RAID controller are the exact same drive, so I’m guessing this was just a bad one in the batch.