Highly available snapshots

A highly available (HA) snapshot is a volume group snapshot that exists on both systems with the same data, and can be restored without suspending high availability (HA) for the partition.

A local snapshot is a volume group snapshot that exists on a single system. Restoring from a local snapshot results in suspension of high availability for the partition until a resynchronization of the restored data is completed.

HA snapshots can be used in any configuration that supports high availability, including as part of a 3-site configuration for high availability and disaster recovery.

If both systems support this feature, HA snapshots are enabled by default when configuring high availability on a partition by using the GUI, or when creating a replication policy for HA by using the CLI or REST API.

On the management GUI, the HA status is displayed on the Snapshots tab of the volume group. On the CLI or REST API, the ha_status field indicates whether a snapshot is HA or local.

You can use the management GUI to enable or disable HA snapshots from the replication tile on the partition. Enabling or disabling HA snapshots affects new snapshots only and any existing snapshots are unchanged. To enable or disable HA snapshots by using the CLI, create a new replication policy with the desired properties and then assign the new replication policy to the partition.

Using HA snapshots

With HA snapshots enabled:
  • Snapshots can be taken on the active management system of an HA partition:
    • If HA is established, HA snapshots are created.
    • If HA is not established, local snapshots are created.
  • The snapshot policy that is associated with the volume group on the active management system defines the scheduling of snapshots.
  • Configuring a snapshot policy on both HA systems ensures that snapshots continue to be triggered if the active management system changes. Safeguarded and retention settings for snapshots are defined by the settings of the volume group on the active management system.

HA snapshots can be removed from the active management system of the HA partition and it results in removal of the snapshots on both systems.

An HA snapshot can be restored from the active management system of the HA partition, either the whole volume group or a single volume:
  • If HA is established, the restore is applied on both systems without loss of HA.
  • If HA is not established, the restore is applied on the active management system and gets synchronized with other system when HA is reestablished.

Clones or thin-clones can be created from an HA snapshot outside of the HA partition on either of the HA systems and are created only on the system where the command is run.

If a partition has HA snapshots, removing HA from that partition results in HA snapshots being converted to local snapshots.

HA thin clone volumes

An HA thin clone volume is a host-accessible copy of a specific volume from a HA snapshot. An HA thin clone volume exists on both systems with the same data.

HA thin clone volumes can be used in any configuration that supports high availability, including as part of a 3-site configuration for high availability and disaster recovery.

An HA thin clone volume can be created on the active management system of an HA partition:
  • If the volume type is thinclone.
  • If the source snapshot is an HA snapshot.
  • If the thin clone volume is placed into a volume group within the same HA partition.

HA thin clone volumes can be refreshed. The refresh is applied on the active management system and gets synchronized with the non active management system when HA is reestablished.

HA thin clone volumes can be converted to clones. The convert to clone happens only on the system where the command was run.