Repairing AWS CVO instance with a failed disk (EBS volume)
Applies to
- Cloud Manager (Service Connector)
- ONTAP 9
- Cloud Volumes ONTAP (CVO)
- Amazon Web Services (AWS)
- AWS Elastic Block Store (EBS) volume
Issue
Note: At times, disk failures occur even in the cloud. The following is a guide to restoring a Cloud Volumes ONTAP HA instance to a healthy state after such a failure.
- CVO generates AutoSupport:
HA Group Notification (FILESYSTEM DISK FAILED Shelf -, Bay -, Model [PHA-DISK ], S/N [xxxxxxxxxxxxxx]) ERROR
HA Group Notification (DISK FAILED Shelf -, Bay -, Model [PHA-DISK ], S/N [xxxxxxxxxxxxxxx]) ERROR
- HA Group Notification (SYNCMIRROR PLEX FAILED) ALERT
- Plex containing the failed disk is
offline
Plex /aggr1/plex1 (offline, failed, inactive, pool1)
RAID group /aggr1/plex1/rg0 (partial, advanced_zoned checksums)
- Aggregate is degraded
RAID Disk Device HA SHELF BAY CHAN Pool Type RPM Used (MB/blks) Phys (MB/blks)
--------- ------ ------------- ---- ---- ---- ----- -------------- --------------
data FAILED N/A 8257528/ -
data 0d.25 0d - - SA:A 1 VMDISK N/A 8257528/16911419264 8388607/17179869176
data 0d.4 0d - - SA:A 1 VMDISK N/A 8257528/16911419264 8388607/17179869176
data 0d.33 0d - - SA:A 1 VMDISK N/A 8257528/16911419264 8388607/17179869176
data 0d.16 0d - - SA:A 1 VMDISK N/A 8257528/16911419264 8388607/17179869176
data 0d.8 0d - - SA:A 1 VMDISK N/A 8257528/16911419264 8388607/17179869176
Raid group is missing 1 disk.
- The EBS volume state changes to "broken"