February 23, 201412 yr I've had two drives that have been set to read only by unRAID in the past 4 days. Both of them after running reiserfsck -check have advised to do a rebuild-tree. Both of the drives are some of my older 2TB drives. Is this just an odd occurrence or is there some underlying problem that is making this happen? EDIT: forgot the "w" in two.
February 24, 201412 yr Author I pulled the first drive and recovered from backup on a different unRAID box. I can pull that SMART report when when I get home. http://pastebin.com/raw.php?i=dyUk919T - Here is the other one. I ran an extended test the Dynamix Health page last night, that passed as well.
February 24, 201412 yr Author That report looks fine. Check the drive cabling. I've been doing some research, wouldn't the CRC errors being much higher if it was a cable issue? They are in hot-swaps trays and everything is secure. I'm going to run the parity-check a couple of times to give my machine a little bit of stress test and read all of the data a couple of times.
February 24, 201412 yr Reiserfsck --rebuild-tree will more than likely fix those drives reiserfsck.pdf reiserfsck_--rebuild-tree_and_when_backup_saves_your_life.pdf
February 25, 201412 yr Author Ya. That's what the check told me to run and I'm back up and going. I'm just concerned that 2 different drives had corruption within 4 days of each other
February 25, 201412 yr 2 in 4 days Could be a coincidence. See here for a story about a guy who was struck by lightning: Texan Man Struck by Lightning Twice on the Same Day I'm not sure what that guys problem was, but did you suffer from power outages when the drives failed? And was your unRAID server doing heavy I/O when the power failed? Is your power supply adequate? (power supplies often don't quite meet their rated Watts so be sure you've some spare power...the torque required at spin ups takes more power than steady state spinning.) You could also run a MEMTEST overnight just to check for a fleeting RAM problem.
February 25, 201412 yr Author 2 in 4 days Could be a coincidence. See here for a story about a guy who was struck by lightning: Texan Man Struck by Lightning Twice on the Same Day I'm not sure what that guys problem was, but did you suffer from power outages when the drives failed? And was your unRAID server doing heavy I/O when the power failed? Is your power supply adequate? (power supplies often don't quite meet their rated Watts so be sure you've some spare power...the torque required at spin ups takes more power than steady state spinning.) You could also run a MEMTEST overnight just to check for a fleeting RAM problem. I would rather have 2 hard drives fail than be struck by lighting twice.... 9 drives/Xeon on a 850W PSU. More info on setup in sig. Also on a UPS to fight brownouts. Memtest may not be a bad idea.. but its ECC as well. I'll run a few passes tonight, thanks for the idea.
February 25, 201412 yr I should'a read your signature...you should have power to spare. Were they both on the same array? (not that I have a solution if they were...just curious about the odds of getting struck twice
February 25, 201412 yr Author I should'a read your signature...you should have power to spare. Were they both on the same array? (not that I have a solution if they were...just curious about the odds of getting struck twice Ya.. both were on my Main array. No worries on the sig. Had to click a link... who wants to do that
February 25, 201412 yr Ya.. both were on my Main array. No worries on the sig. Had to click a link... who wants to do that
Archived
This topic is now archived and is closed to further replies.