Disaster Recovery Management

[ChakshuGautam]

• Define DR Management strategy for this.
• Implementation.

References:

1. https://github.com/EnterpriseDB/barman
2. https://github.com/pgbackrest/pgbackrest

[PiyushRaj927]

Hi @ChakshuGautam, could you please provide more information on this issue? Thank you!

[nitishmalang]

Hi @ChakshuGautam, could you please provide more information on this issue? Thank you!

yes @ChakshuGautam please

[PiyushRaj927]

@ChakshuGautam, I would like your input on this suggestion. My suggestion is to use pgBarman (https://pgbarman.org/) in a separate container alongside our PostgreSQL installation. It enables Point-in-Time Recovery (PITR) for efficient data restoration during emergencies. Do you think this solution meets our requirements?

cc: @jayanth-kumar-morem @singhalkarun

[ChakshuGautam]

@PiyushRaj927 This is exactly what I was looking for. We have used Barman before and it works as expected. The wal logs are a bit too much in size so look for compression.

[jayanth-kumar-morem]

Hey @ChakshuGautam ,

PGSQL has built-in support for compressing WAL.
We can control this using the wal_compression configuration parameter in the postgresql.conf file. When this parameter is set to on, PostgreSQL will try to compress WAL data when a full page image is written to the WAL. This can significantly reduce the amount of WAL data.

[jayanth-kumar-morem]

@PiyushRaj927 Have you started working on this? If not I can take this one up.

[PiyushRaj927]

No @jayanth-kumar-morem, I have not started working on the issue.

[jayanth-kumar-morem]

Alright, I'll get on this