Making checkpoint_timeout longer to prevent freqent checkpointing
We are planning our PostgreSQL database to move to Azure Database for PostgreSQL.
Now, we are working on performance analysis and tuning (for database refactoring) prior to moving to Azure, and we have found that frequent checkpoints are generating heavy load.
So, we have checked the checkpoint parameters on Azure, and we afraid of that checkpoint_timeout (5min) is too short and it couldn't be changed by the user. It could generates frequent checkpoints.
I think 30~60min for the checkpoint_timeout parameter is very common , so I hope Azure would have that as the default value.
 Basics of Tuning Checkpoints