"PostgreSQL Bugs List" <pgsql-bugs@postgresql.org> writes:
> My App is doing 500 inserts initially in each table. After all this done, we
> inserting 50 in each table and deleting previous 50 records every seconds.
> System performs well for awhile (<30 Hrs). After 30 hrs I seen that dir size
> of $PGDATA/base dir is keep on growing and to goes up to 2G in 48 hrs.
Is it the tables that are growing, or the indexes? (If you're not sure,
look at pg_class.relpages, or the output of vacuum verbose, to see which
files are getting out of hand.)
If it's the tables then you probably need to increase your FSM settings.
If it's the indexes then you're kind of stuck --- 7.3 doesn't have any
solution to index bloat other than periodic REINDEX, which may be too
much of a performance hit for you.
> Every time vacuum is triggered, the system goes extreamly slugginsh,
> and results in various errors like deadlock detected(confirmed in the
> $PGDATA/../LOG/logfile).
Plain vacuums shouldn't result in any deadlocks. Are you doing anything
unusual like taking out explicit exclusive locks on tables?
> ++++++++++++++++++++++++++++++++++++++++++++++++++ NOTE: I can not use
> Postgres 7.4 and higher releases beacuse postmaster crashes gauranteed
> in (20hrs). I have already reported this bug many times (Bug # 1104 is
> one of them).
Indeed, and you have not yet provided any information with which anyone
else could reproduce or track down the problem. We're not ignoring you,
we just can't do anything without better information.
regards, tom lane
