Symptoms

Storage Pool reached 100% full. A storage process has panicked when a pool became full and one or more File Systems (FS) or Logical Unit Numbers (LUN) show as offline and requiring recovery. Any snapshots for one or more FS or LUN which are marked as offline need recovery as well.

Cause

The sequence of the panic: The pool becomes 100% full, and the array starts invalidating the snapshots.The array starts to unmount the snapshots.The array tries to update the metadata and superblock in order to allocate getSlice. This fails due to the pool being 100% full.The operation to unmount the snapshots timeout and cause an SP Panic.

Resolution

The FS, LUNs, and snapshots need free space in the pool in order for a recovery attempt to be started. This is done by expanding the impacted pool. If the impacted pool cannot be expanded, then free space must be made in the pool by one of the following methods: Delete snapshotsExpand the storage pool (using existing drives or purchasing additional drives if there are slots available)Remove unneeded resources (LUNs or FSs)Run the space reclaim procedure (from the host-side) to potentially free-up space which has not been zeroed-out on the array side Reclaiming VMFS deleted blocks on Thin Provisioned LUNs (external link) Investigate preallocated space concerns Dell Unity: Pool Space Issues Due To Large Levels Of Preallocated Space If further help is needed, contact Dell Technical Support and reference this Dell Knowledgebase Article.

Support Cases