Hi,
We see in one of the RDS postgres instances, from pg_stat_user_tables , the auto vacuum and auto analyze happening on the tables without our manual intervention.
So is auto vacuum analyze is sufficient to make sure optimal stats and unbloated table structure in place or should we do it manually based on the type of tables like
for e.g.
if a table is having just insert only types and getting similar amount of data throughout the day 24*7
VS
a table which is getting bulk data load once a day only
VS
a volatile table with truncate load kind of operation
VS
a table with heavy Update/deletes also happening along with inserts throughout the day 24*7.
Will auto vacuum/analyze take care of all such tables, or we need to do it manually in certain scenarios?
And if any suboptimal plan is taken by the optimizer for a specific query suddenly, because of the missing stats , how to catch that. Basically, how to get the old plan hash and get it compared with the current plan hash?
Regards
Veem