misterbeetz Posted August 19, 2013 Share Posted August 19, 2013 Hello all. I noticed something peculiar with my machine and I'm wondering if anyone can tell me if this is normal or not... So I just upgraded to version 5.0-rc16c from 4.7 due to the fact that I really need to use some 3TB drives I had lying around. Once I upgraded I checked that the server was working correctly and it seemed fine (before the upgrade the drives seemed to be fine as well: no read errors etc.). Next I went ahead with upgrading my parity drive from a 2TB to a 3TB. Once that started happening though I noticed that I was slowly getting read errors on one of my data drives. When the upgrade finished I was left with 117 read errors on said data drive. The web interface however now says that the parity is valid... So here is my question. My understanding was that unraid would normally correct read errors on the fly because it had parity data available to check against however how is this possible during a parity drive upgrade when parity is not yet complete/available? Can this mean that the new parity information might not actually be valid? Any clarification on this would be much appreciated. I have more details on the data drive below. - The error count for my disk 3 as per the web interface is 117 (the other drives seem fine) - my syslog has multiples of these errors: Aug 18 14:00:32 Tower kernel: md: disk3 read error, sector=1851646416 Aug 18 14:00:32 Tower kernel: md: multiple disk errors, sector=1851646416 - Reallocated_Sector_Ct is 0 - Current_Pending_Sector is 7 - UDMA_CRC_Error_Count is 0 I have attached some of my syslog showing the errors in addition to the smart report for the suspect data drive. Many thanks in advance! smart-sde_8-18-2013.txt syslog_8-18-2013.txt Quote Link to comment
JonathanM Posted August 20, 2013 Share Posted August 20, 2013 So I just upgraded to version 5.0-rc16c from 4.7 due to the fact that I really need to use some 3TB drives I had lying around. Once I upgraded I checked that the server was working correctly and it seemed fine (before the upgrade the drives seemed to be fine as well: no read errors etc.). Next I went ahead with upgrading my parity drive from a 2TB to a 3TB. Did the parity check you did after the software upgrade and before you switched the parity drive complete normally? Quote Link to comment
misterbeetz Posted August 20, 2013 Author Share Posted August 20, 2013 Did the parity check you did after the software upgrade and before you switched the parity drive complete normally? Unfortunately I did not do any manual parity checks specifically for this occasion (I was too impatient ). I do however have automatic parity checks set to run at the beginning of each month and there were no noticeable problems back then. Ah boo, hindsight is always 20/20... Quote Link to comment
dgaschk Posted August 20, 2013 Share Posted August 20, 2013 Put the old parity drive back and then rebuild the error disk. Do a parity check and then proceed. Quote Link to comment
stealth82 Posted August 23, 2013 Share Posted August 23, 2013 Same thing occurred to me. But my operation was nastier. I swapped the parity disk with a data disk of the array I had previously emptied. Started a new config and the parity check is still in progress. 102 errors so far. Should I started getting worried? Quote Link to comment
dgaschk Posted August 23, 2013 Share Posted August 23, 2013 Any operation that reads from disk3 will fail. Parity cannot be built correctly if one of the disks can not be read correctly. Disk 3 needs to be rebuilt before anything else. Replace the original parity drive and set a new config (Utils-New Config). Check the box that indicates that parity is correct and start the array. Stop the array and replace disk 3. Start the array and rebuild disk 3. Then run a parity check. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.