This guide can be used to recover from one or more volume snapshots created on FlashArray for a PostgreSQL cluster. Recovery can be done to the original system from where the volume snapshot was taken or to a different system (cloning).
Note:
This guide assumes that a storage snapshot of a PostgreSQL cluster has been taken as per Creating Volume Snapshots of PostgreSQL on FlashArray.
The method for recovering a PostgreSQL cluster from volume snapshots on FlashArray will vary under the following scenarios:
- The volume snapshots are created on a system using ther_volume_and_file_system_architectural_layout_02.html#r_volume_and_file_system_architectural_layout_02__r_volume_and_file_system_architectural_layout_02-tbl1 or the cluster is initialized at a different location.
- The cluster components are located on separate volumes.
- There are multiple tablespaces in the cluster deployed on multiple volumes.
Note:
The recovery method for application consistency is the same as crash consistency. The difference between the two approaches is how consistency is guaranteed.