On our database of about 5GB we vaccuum all of our 12 tables (only one
is huge, all others have about 100,000 rows or so) every hour or so.
But we also have autovaccuum enabled. Is this okay? Do the two vaccuum
processes contradict each other, or add unnecessary load to the
system?
The reason we
On Thu, Aug 28, 2008 at 07:53:17PM +0800, Phoenix Kiula wrote:
On our database of about 5GB we vaccuum all of our 12 tables (only one
is huge, all others have about 100,000 rows or so) every hour or so.
But we also have autovaccuum enabled. Is this okay? Do the two vaccuum
processes
On Thu, 2008-08-28 at 19:53 +0800, Phoenix Kiula wrote:
On our database of about 5GB we vaccuum all of our 12 tables (only one
is huge, all others have about 100,000 rows or so) every hour or so.
if you refer to manual VACUUM or VACUUM FULL every hour is probably too
much. You should aim your
On Thu, Aug 28, 2008 at 9:47 AM, Joao Ferreira gmail
[EMAIL PROTECTED] wrote:
http://www.postgresql.org/docs/8.3/static/maintenance.html
you'll find that once in a while (start at once/week and build up or
down from there) you can/should:
- vacuum full
- reindex your tables
- reindex your
In response to Joao Ferreira gmail [EMAIL PROTECTED]:
On Thu, 2008-08-28 at 19:53 +0800, Phoenix Kiula wrote:
On our database of about 5GB we vaccuum all of our 12 tables (only one
is huge, all others have about 100,000 rows or so) every hour or so.
if you refer to manual VACUUM or
For the record:
Bill Moran escribió:
The naptime at 600 is probably a bad idea. If you only have one user
database on this system, then it only gets investigated by autovac once
every 40 minutes (template0 ... template1 ... postgres ... yourdb)
Consider that autovac uses very little
In response to Alvaro Herrera [EMAIL PROTECTED]:
For the record:
Bill Moran escribió:
The naptime at 600 is probably a bad idea. If you only have one user
database on this system, then it only gets investigated by autovac once
every 40 minutes (template0 ... template1 ... postgres
Hello,
I'dd like to apologise about my short knowledge of VACUUM FULL and
REINDEX.
I'm just stating what I do in my case. I don not know if it is a corner
case or not.
I've been dealing with this specific application which is very demanding
for Postgres for about 2 years.
When autovacuum was
On Thu, Aug 28, 2008 at 05:15:06PM +0100, Joao Ferreira gmail wrote:
When autovacuum was introduced, I kept the weekly VACUUM FULL because it
efectively brings disk ocupatio down, dispite it grows back after a few
hours. It's just re-assuring to me to make sure that at least one of the
In response to Joao Ferreira gmail [EMAIL PROTECTED]:
When autovacuum was introduced, I kept the weekly VACUUM FULL because it
effectively brings disk occupation down, despite it grows back after a few
hours. It's just re-assuring to me to make sure that at least one of the
vacuums it's
10 matches
Mail list logo
