CONTAP-377578: SAS drive reporting not ready status may cause high client IO latency
Issue
If a SAS drive returns an IO with not ready status then ONTAP keeps retrying the IO to wait for the drive to become ready. In some cases it may take the drive too long to become ready which can lead to long latencies for client IOs. This happens because ONTAP is not temporarily offlining the drive from the RAID group when this IO status happens.
EMS log:[?] Mon Oct 21 08:38:25 +0900 [node1: scsi_cmdblk_strthr_admin: scsi.cmd.notReadyCondition:notice]: Disk device 0d.16.1: Device returns not yet ready: CDB 0x28:628b3180:0008: Sense Data SCSI:not ready - Drive spinning up (0x2 - 0x4 0x1 0x0)(5970).
Note:2/4/1(0x2 - 0x4 0x1 0x0) is a standard SCSI check condition. It means LOGICAL UNIT IS IN PROCESS OF BECOMING READY.
This might also cause long CPs as observed in the EMS log:Mon Oct 21 08:40:21 +0900 [node1: wafl_exempt07: wafl.cp.toolong:error]: Aggregate aggr1 experienced a long CP.