israel.garcia@cimex.com.cu wrote:
I have a server running some dababases in CentOS4.5 fully updated, but this morning I read this error in messages log:
Trover st0:Block limits 1- 16777215 bytes
172.25.8.236: Servinaves ATA ABNORMAL DTATUS 0X20DF ATA ABNORMAL DTATUS 0X20DF ATA ABNORMAL DTATUS 0X20DF ata comand 0x35 timeout, stat 0xd0 host - stat 0x21 atat1: status = 0xd0 Busy SCSI error: 2000 returns code0x8000002 current sda sense key aborted command additional sense:Scsi parity error end- request:110 error, devsda,sector 87008471
I have mdadm running RAID1 with two disk!
Well you used to have 2 disks... I suspect that one of them was kicked out for this error. Does 'cat /proc/mdstat' show [UU] (both active) or an '_' in one of the positions?