September 27, 201213 yr Yesterday a disk red-balled and showed disabled. I reassigned the same disk and it was rebuilt. After the rebuilt I started a non-corrective parity-check resulting in 1863 parity errors. But mymain in unmenu shows Parity updated 1863 times to address sync errors. Should this happen with a non-corrective parity sync? What to do now? The parity check started on 26 sep at 23:16:xx. Trouble began a while later when I was asleep. Please see the attached syslog. syslog-2012-09-27.txt
September 27, 201213 yr The message is incorrect. Parity was NOT updated if you selected a non-correcting check. This is a known issue. Attach a SMART report for disk1.
September 28, 201213 yr Author Here are the smart reports of the other data-disks. I can't pull a smart report of the parity and cache disk because they are RAID volumes on a ARC1200 controller. Could these sync errors have been caused by inserting 2 disks in a CSE-M35T while the server was running and rebooting the server afterwards while SABNZB was downloading and after the reboot 2 preclears were started in screen on the two newly inserted disks. The parity sync was then started too. Maybe I was asking too much? I also saw parity sync errors at 1465384960 and higher, what does this number say? How much data does 1863 sync errors mean? disk2.zip disk3.zip
September 28, 201213 yr Author I connected the 2 disks on the ARC1200 straight to the motherboard and pulled the smart reports from them which you find here. wd1601.zip wd5852.zip
September 29, 201213 yr Author Will do that later(coming night). A parity sync just finished. Immediately upon starting a parity check another disk red-balled. I tried to re-assign the same disk but then it showed as 750 GB instead of 1 TB resulting in "replacement disk too small". After a reboot it was ok. Maybe I have been fooling around too much with the NCQ setting in Disk settings. Or is there some compatibility problem with the M1015 card? I tried 5.0RC-3 and now 5.0b12a. Luckily this is my test server.
September 29, 201213 yr What model PSU and how many disks? Use version rc8a. There have been no issues reported this version.
September 29, 201213 yr Author I have SM X9SCM-F, E3-1240, 16 GB ECC memory. For future ESXi build Seasonic X760 PSU 2 disks on Areca ARC1200 3 data disks on a M1015 controller + 3Ware forward breakout cable Will grow to 15-20 data disks when all trouble is gone. data disks are in SM CSE-M35T1 cages.
October 3, 201213 yr Author Disk2 redballed again yesterday evening. It is a 1 TB Samsung. I stopped the array and then started it again. The disk then showed as 750 GB. The disk is on a flashed M1015 controller and unraid 5beta12a is used. For disks on a M1015, should the spindown times be set in unraid or is this controlled by the controller itself? I have attached the latest syslog syslog-20121003-000033.zip
Archived
This topic is now archived and is closed to further replies.