From: | Lists <lists(at)benjamindsmith(dot)com> |
---|---|
To: | Steve Crawford <scrawford(at)pinpointresearch(dot)com> |
Cc: | pgsql <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Unexpectedly high disk space usage RESOLVED (Manual reindex/vacuum) |
Date: | 2012-11-12 04:17:39 |
Message-ID: | 50A07863.1020007@benjamindsmith.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 11/09/2012 05:26 PM, Steve Crawford wrote:
> Bloat in pg_attribute would correlate with A) (or any constant
> creation/destruction of tables). You can vacuum and/or reindex the
> system tables if you are connected as the superuser but you are better
> off preventing bloat by appropriate adjustment of your configuration
> settings. However note that if you do frequent bulk
> creation/destruction of tables you could end up bloating the attribute
> table between vacuum runs and may need to periodically manually shrink
> it.
>
Steve,
Our system divides customers into distinct databases, however customers
are often clustered. (Think: different locations of 7/11) and so we have
to aggregate data from different databases. We do this with dblink to
get the data and temp tables to collate it, which appears to be a cause
of the bloat we're seeing.
-Ben
From | Date | Subject | |
---|---|---|---|
Next Message | Thalis Kalfigkopoulos | 2012-11-12 04:18:00 | VACUUM details (vacuum_freeze_min_age/vacuum_freeze_table_age) |
Previous Message | Craig Ringer | 2012-11-12 03:23:22 | Re: Can dml realize the partition table's rule and make good execution plan? |