Ubuntu raid 1 write errors
Posted
by
Micah
on Server Fault
See other posts from Server Fault
or by Micah
Published on 2011-03-18T02:46:51Z
Indexed on
2011/06/27
0:24 UTC
Read the original article
Hit count: 521
I have an Ubuntu server set up with two SATA drives in a RAID 1 configuration with MDADM. The machine is used to record raw video, which involves a lot of writing to the disk. Sometimes during video recording the computer will crash, will the following errors in kern.log:
Mar 15 10:39:41 video kernel: [414501.629864] ata2.00: exception Emask 0x10 SAct 0x0 SErr 0x400100 action 0x6
Mar 15 10:39:41 video kernel: [414501.629870] ata2.00: BMDMA stat 0x26
Mar 15 10:39:41 video kernel: [414501.629875] ata2.00: SError: { UnrecovData Handshk }
Mar 15 10:39:41 video kernel: [414501.629880] ata2.00: failed command: WRITE DMA EXT
Mar 15 10:39:41 video kernel: [414501.629889] ata2.00: cmd 35/00:00:28:6d:f6/00:04:06:00:00/e0 tag 0 dma 524288 out
Mar 15 10:39:41 video kernel: [414501.629891] res 51/84:b1:77:6e:f6/84:02:06:00:00/e0 Emask 0x30 (host bus error)
Mar 15 10:39:41 video kernel: [414501.629896] ata2.00: status: { DRDY ERR }
Mar 15 10:39:41 video kernel: [414501.629899] ata2.00: error: { ICRC ABRT }
Mar 15 10:39:41 video kernel: [414501.629910] ata2.00: hard resetting link
Mar 15 10:39:41 video kernel: [414501.973009] ata2.01: hard resetting link
Mar 15 10:39:41 video kernel: [414502.482642] ata2.00: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
Mar 15 10:39:41 video kernel: [414502.482658] ata2.01: SATA link down (SStatus 0 SControl 300)
Mar 15 10:39:41 video kernel: [414502.546160] ata2.00: configured for UDMA/133
Mar 15 10:39:41 video kernel: [414502.546203] ata2: EH complete
Is this the result of faulty drives? Is software RAID just not performant enough for data rates ~15 MB/s, even with a quad-core i7?
Thanks for your help.
Edit: cat /proc/mdstat
returns this:
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md0 : active raid1 sdb1[1] sda1[0]
976760768 blocks [2/2] [UU]
unused devices: <none>
© Server Fault or respective owner