Hi All Thanks for all your valuable inputs,
Here is some more data, Though we have 150 GB free space spread across 500 Tables , Every alternative day DB is growing with 1 GB rate. Also,We have manual vacuum job scheduled to run weekly basis, So seems to be space is not reusing all the time ? So conclude the requirement here , The only way to parallelism is multiple script. And no need to do REINDEX exclusively. Question : Do we need to consider Table dependencies while preparing script in order to avoid table locks during vacuum full ? At present Maintenance work memory set to 20 GB. Question : Do we need to tweak any other parameters ? Note: We are planning this activity with Application Downtime only. Let me know if i missed anything. Regards, Raj On Wed, Apr 3, 2019 at 8:42 AM rihad <ri...@mail.ru> wrote: > > And future updates can reuse it, too (an update is very similar to an > > insert+delete). > > > Hm, then it's strange our DB takes 6 times as much space compared to > freshly restored one (only public schema is considered). > > > Not if autovacuum has a chance to run between updates. > > Ours is run regularly, although we had to tweak it down not to interfere > with normal database activity, so it takes several hours each run on the > table. We did that by setting autovacuum_vacuum_scale_factor = 0.05 from > default 0.2. > > >
