May 24, 201610 yr Just completed my first parity check since moving to 6.2b21 and it completed with 12 errors, can someone have a look at my diags and let me know what happened? https://www.dropbox.com/s/hkri20bhkgio2yk/tower-diagnostics-20160523-2023.zip?dl=0
May 24, 201610 yr Community Expert Many UDMA_CRC errors on disk20, these would noticeably slowdown the parity check, replace SATA cable and monitor this value: Device Model: ST4000DM000-1F2168 Serial Number: Z303SBXG 199 UDMA_CRC_Error_Count 0x003e 200 200 000 Old_age Always - 4553 A value increase of 2 or more meens there's still a problem. Besides the CRC errors, looks like there were also some other errors on this disk that are next to parity corrections, so do another check after replacing the cable.
May 24, 201610 yr Author Oddly enough UnRaid kicked off another parity check all by itself about 8hrs ago, I haven't touched any cables, so we'll see what happens. Thanks
May 24, 201610 yr Community Expert With so many UDMA_CRC errors I would cancel it, replace the cable and start another. Also parity check only starts by itself if scheduled or after an unclean shutdown, e.g., the server rebooted itself.
May 24, 201610 yr Author Yeah it was scheduled, and I still see lots of CRC errors in the log too. Which disk has the bad cable so I know which one to replace or fix?
May 24, 201610 yr Author Sorry wasn't paying attention. I powered the server down, checked all the cables. I did recently buy a SUPERMICRO AOC-SAS2LP-MV8 PCI-Express 2.0 x8 SATA / SAS 8-Port Controller Card but I am using my own breakout cables that I got a few years ago with another card, so I guess its possible one of the two breakout cables has issues and may need to be replaced, I will monitor the drive and start another parity check, thanks.
May 24, 201610 yr And don't tie up any cabling. The most reliable cabling (when in doubt as to the quality of its shielding) is laying within the case completely random
May 25, 201610 yr Author Ok so another parity check ran, this time returning 5 sync errors, logs attached, can anyone tell me what it found this time? tower-diagnostics-20160525-1323.zip
May 25, 201610 yr Community Expert Everything looks good, took a lot less time also, 28 hours instead of 43, I suspect the previous run did some wrong corrections because of the errors, next check should find 0 sync errors.
Archived
This topic is now archived and is closed to further replies.