I've setup a master / slave Postgresql (13.4) group, 1 master and 1 slave, using WAL warm standby.
(refer: postgresql.org/docs/current/warm-standby.html#STANDBY-SERVER-SETUP)

The slave gone out of sync, and can't recover due to WAL not exists anymore on master, refer another question for details: Postgresql - recover slave after out of sync with master using WAL

Now, I'am re-doing the slave. There are 3.7T data on master, and it's increasing pretty quickly, roughly 30Gb per day ?

To avoid slave go out of sync again, I'm trying to increase the WAL retention.

Here is what I plan to do:

  • max_wal_size = 20Gb
  • min_wal_size = 10GB


  • A-1. Are the above param values proper?
  • A-2. Does increasing these param has a bad side effect ?
  • A-3. I only need to apply these 2 params to master, right?
  • B. Is there more parameter that I should config?
  • C. Any better way to avoid slave out of sync ?

Thank you.

Source link