| From: | Wes <wespvp(at)syntegra(dot)com> |
|---|---|
| To: | Postgresql-General <pgsql-general(at)postgresql(dot)org> |
| Subject: | Vacuum time degrading |
| Date: | 2005-02-28 23:02:04 |
| Message-ID: | BE48FF0C.7C01%wespvp@syntegra.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general pgsql-hackers |
We've been doing a vacuum every night. With about 254 million rows, it was
taking abut 0.8 hours. Now a few months later with about 456 million rows
it is taking 4.8 hours. I did check the archives and didn't find anything
like this.
Why is the vacuum time not going up linearly? Since vacuum has to
sequentially read the entire database, I would have expected a linear
increase - about 1.5 hours now.
There are currently no deletes or modifies to the database - only inserts.
This is on PostgreSQL 7.4.5, RedHat ES 3.0.
Wes
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Frank van Vugt | 2005-02-28 23:07:48 | invalid multibyte character for locale |
| Previous Message | Ron Mayer | 2005-02-28 23:00:53 | vacuum_cost_delay & VACUUM holding locks on GIST indexes |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Nicolai Tufar | 2005-02-28 23:22:24 | Thread-safe snprintf() vsnprintf() and printf() |
| Previous Message | Bruce Momjian | 2005-02-28 22:41:26 | Re: [PATCHES] snprintf causes regression tests to fail |