HI Julien, In my current database index size is very large compare than table (if table size is 30 GB and index size is 90GB) , that why We running full vacuum monthly basis to increase the system performance and data reclaim . Auto Vacuum is also configuration in system. As per your suggesion if we increase " maintenance_work_mem " paramater size to 1GB this will be help to improve full vacuum performance?. System configuration as below: CPU - 42 RAM -70GB Regards, Ram Pratap. -----Original Message----- From: Julien Rouhaud [mailto:rjuju123@xxxxxxxxx] Sent: 08 February 2022 12:38 To: Ram Pratap Maurya Cc: 'pgsql-admin@xxxxxxxxxxxxxxxxxxxx'; Ashish Chugh Subject: Re: PostgreSQL Full Vacuum Taking 5 to 6 hrs. Hi, Please don't top post on those forums (https://wiki.postgresql.org/wiki/Mailing_Lists#Email_etiquette_mechanics) On Tue, Feb 08, 2022 at 06:59:07AM +0000, Ram Pratap Maurya wrote: > > we are running VACUUM FULL on all table and its take 5 to 6 Hrs. The only thing that can be done to make it faster is to use a 1GB maintenance_work_mem. But why do you need to do frequent database wide VACUUM FULL in the first place? Do you really have massive bloat on every single table in the database? Do you have adequate autovacuum configuration, or additional VACUUM jobs if that's not enough?
