Known limitations for WarehousePG Disaster Recovery

These are the current limitations of WarehousePG Disaster Recovery. Where applicable, workarounds are included to help mitigate the impact.

Concurrent restore-point creation from multiple hosts isn't supported. Running whpg-dr create-restore-point concurrently from different machines against the same cluster produces unreliable restore-point records in the backup repository. Run restore-point creation from one designated host at a time.
The DR cluster must not be initialized before restore. Running gpinitsystem on the DR cluster before whpg-dr restore causes the restore to fail. Follow the WarehousePG install guide through Creating the Data Storage Areas, then stop. Don't proceed to Initializing WarehousePG. If gpinitsystem has already been run, delete the data directories on all DR hosts before attempting a restore.
No pre-restore WAL completeness check. whpg-dr restore doesn't verify that all WAL files required to reach the target restore point are present in storage before starting. If a WAL file is missing, the restore fails part-way through:
```
WARNING: WAL replay did not complete. Data directories for 1 coordinator and N segments
were modified, but the restore-last state file has NOT been updated.
```
The failing segment's startup.log shows a pg_ctl start failure but doesn't identify which WAL file is absent. Run whpg-dr check and verify the storage repository before restoring. If the restore fails, re-run it with --resume-failed after investigating log output in $HOME/gpAdminLogs.
No failback automation. After promoting the DR cluster to a writable primary, reversing roles isn't automated. To use the original primary as a DR site, re-initialize it as a new DR cluster and run a fresh base backup.