Disks failing in a stretch MetroCluster FC due to multiple scsi.cmd errors
Applies to
- ONTAP 9
- Stretched MetroCluster FC
- ATTO FibreBridges
Issue
- ONTAP reporting multiple
scsi.cmd
and high latency errors for disks:
Mon Apr 03 05:57:53 +0100 [JIFN1001: slifc_intrd: scsi.cmd.checkCondition:error]: Disk device 3a.125L13: Check Condition: CDB 0x28:015b3800:0200: Sense Data SCSI:aborted command - (0xb - 0x4b 0x4 0x50)(2).
Mon Apr 03 05:59:04 +0100 [JIFN1001: disk_latency_monitor: shm.threshold.highIOLatency:error]: Disk 3a.125L13 exceeds the average IO latency threshold and will be recommended for failure.
- The disks are put in maintenance and eventually failed
- All
scsi.cmd
errors are pointing to a specific SAS port on an ATTO bridge - Errors contunue even after the failed disks are replaced