Dell EMC Unity: Block/File storage resources invalidation behavior for over subscribed pools

The snapshot retention policy generally aims to reduce the likelihood that an oversubscribed pool becomes entirely full as the result of snapshots, this condition can occur in rare cases when snapshots exist that are not configured for automatic deletion or are retained for too long. Because snapshots configured for time based automatic deletion or no automatic deletion are not subject to automatic deletion on the basis of the pool or snapshot space thresholds, it is possible that an over- provisioned pool can become entirely filled while at the same time, advertising available space to its clients. In this case, either the host I/O must be failed, or snapshots must be invalidated to make room for new writes. The snapshot invalidation behavior depends on the resource type and configuration.

For block, this behavior is different for Thin LUNs and Thick LUNs.

or file, this behavior depends on the  poolFullPolicy setting, which can be configured using the UEMCLI /stor/prov/fs command.

This provides the administrator with the option to configure each individual file system to either delete all associated snapshots (including system replication snapshots) or fail host I/O to that resource. The default option is failWrites.