MetroCluster backend LUNs show disk failed during E-Series FW update
Applies to
- FAS8200
- ONTAP 9.3
- Fabric Metrocluster
- E-Series Backend
Issue
- While the customer was performing firmware updates on their E-Series storage arrays the attached ONTAP Fabric Metrocluster reported multiple disks failed due to SCSI timeout errors:
Fri Dec 04 20:09:18 EST [node1: config_thread: raid.config.filesystem.disk.not.responding:notice]: File system Disk /node1/plex0/rg0/switch1:12.126L119 Shelf - Bay - [NETAPP INF-01-00 0842] S/N [xxxxxxxxxxxxxxxxxxxxxx] UID [xxxxxxxx:xxxxxxxx:xxxxxxxx:xxxxxxxx:00000000:00000000:00000000:00000000:00000000:00000000] is not responding.
- These errors lead to multiple syncmirror plex failures on both clusters in the Metrocluster.
Example:
Aggregate aggrname (online, raid0, mirror degraded) (block checksums)
Plex /aggrname/plex0 (offline, failed, inactive)
RAID group /aggrname/plex0/rg0 (partial, block checksums)
RAID Disk Device HA SHELF BAY CHAN Pool Type RPM Used (MB/blks) Phys (MB/blks)
--------- ------ ------------- ---- ---- ---- ----- -------------- --------------
data FAILED N/A 13972000/ -
data switch1:12.126L78 0f - - 0 LUN N/A 13972000/28614656000 14000000/28672000000
data FAILED N/A 13972000/ -
data switch1:12.126L82 0f - - 0 LUN N/A 13972000/28614656000 14000000/28672000000
data switch1:12.126L80 0f - - 0 LUN N/A 13972000/28614656000 14000000/28672000000
data FAILED N/A 13972000/ -
Raid group is missing 3 disks.