June 23, 201214 yr Hi guys, I seem to be having major issues with the rebuild of a new disk. It was in the middle of a disk replace (#1) when the power got cut off. I restarted the server (and disk replace #2) and now I woke up this morning to (see attached picture)... It looks like are write errors 134774640 and counting) on disk 2,3 & 5 and all files on these disks are missing. I don't want to halt the current rebuild but is the data on these lost? Or can I wait for the disk replace to complete and rebuild parity so I can get my files back? here is a snippet of the log Jun 23 07:25:37 Tower2 kernel: md: disk2 read error Jun 23 07:25:37 Tower2 kernel: handle_stripe read error: 65088296/2, count: 1 Jun 23 07:25:37 Tower2 kernel: md: disk5 read error Jun 23 07:25:37 Tower2 kernel: handle_stripe read error: 65088296/5, count: 1 Jun 23 07:25:37 Tower2 kernel: md: disk1 read error Jun 23 07:25:37 Tower2 kernel: handle_stripe read error: 65088304/1, count: 1 Help/suggestions would be appreciated!
June 23, 201214 yr Author OK I Stopped the Rebuild at 31% Rebooted and Started (w/o rebuild) I am going to give it another go but if I get more errors like above - does anyone have any steps to solve it?
June 24, 201214 yr I didn't look at the syslog. but inadequate PSU is the very first thing that comes to mind when you have multiple errors on multiple drives at the same time during a rebuild. There are always other reasons like a bad card or disk.. but power is often overlooked.
June 24, 201214 yr Author Ok thanks. Its a fairly new PSU but I will take a look. The second time completed and now it looks like both the Parity and #2 (no files again) have errors (different drives that previous attempts!) so maybe it is the power. It is currently attempting to correct errors by writing corrected parity. hope i get no errors and my files back.
Archived
This topic is now archived and is closed to further replies.