Nice article, sneak peak into the actual WAL file was interesting.
Any idea how often these WAL files are rebased or purged ? Since over years these could build up like anything especially on a write heavy postgres instance.
WAL files could take up huge space onto the disk. That's why they are purged when checkpointing is done. Checkpointing refers to the process when the corresponding statements have been applied to database and persisted successfully and thus post that we don't need those set of statements from WAL file for future use case.
Now iam wondering how database recovery happens during a crash using these checkpoints. May be i should read and understand from postgres official. Btw, very good article that instigates a thought to understand more.
Thank you for this article, was really insightful!
Thanks for reading & appreciating:)
Thanks a lot for this. It contained a lot from scratch to advanced after coming from your previous article of WAL.
Nice article, sneak peak into the actual WAL file was interesting.
Any idea how often these WAL files are rebased or purged ? Since over years these could build up like anything especially on a write heavy postgres instance.
you're correct.
WAL files could take up huge space onto the disk. That's why they are purged when checkpointing is done. Checkpointing refers to the process when the corresponding statements have been applied to database and persisted successfully and thus post that we don't need those set of statements from WAL file for future use case.
Sounds interesting,
Now iam wondering how database recovery happens during a crash using these checkpoints. May be i should read and understand from postgres official. Btw, very good article that instigates a thought to understand more.