плохой диск в зеркале
Hi, All!
Нарисовалась у меня проблемма на боевом сервере.
Периодически вываливается из зеркала один диск:
# cat /proc/mdstat
Personalities : [raid1]
md1 : active raid1 sda1[2](F) sdb1[1]
732571904 blocks [2/1] [_U]
md0 : active raid1 sdd1[0] sdc1[1]
7815488 blocks [2/2] [UU]
Удалось накопать следующее:
Oct 9 10:50:44 bagdo kernel: [4620479.650980] ata1.00: exception Emask 0x10 SAct 0x0 SErr 0x10000 action 0xe frozen
Oct 9 10:50:44 bagdo kernel: [4620479.650980] ata1: SError: { PHYRdyChg }
Oct 9 10:50:44 bagdo kernel: [4620479.650980] ata1.00: cmd 25/00:08:57:10:0c/00:00:27:00:00/e0 tag 0 dma 4096 in
Oct 9 10:50:44 bagdo kernel: [4620479.650980] res d0/d0:d0:d0:d0:d0/d0:d0:d0:d0:d0/c0 Emask 0x12 (ATA bus error)
Oct 9 10:50:44 bagdo kernel: [4620479.659202] ata1.00: status: { Busy }
Oct 9 10:50:44 bagdo kernel: [4620479.659202] ata1.00: error: { ICRC UNC IDNF }
Oct 9 10:50:44 bagdo kernel: [4620479.659202] ata1: hard resetting link
Oct 9 10:50:49 bagdo kernel: [4620484.840165] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
Oct 9 10:50:49 bagdo kernel: [4620484.873459] ata1.00: configured for UDMA/33
Oct 9 10:50:49 bagdo kernel: [4620484.873496] ata1: EH complete
Oct 9 10:50:49 bagdo kernel: [4620484.894862] sd 0:0:0:0: [sda] 1465149168 512-byte hardware sectors (750156 MB)
Oct 9 10:50:49 bagdo kernel: [4620484.900779] sd 0:0:0:0: [sda] Write Protect is off
Oct 9 10:50:49 bagdo kernel: [4620484.900779] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
Oct 9 10:50:49 bagdo kernel: [4620484.924935] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
вопрос, собственно, какую порекомендуете последовательность проверок/замен?
--
DamirX
Reply to: