On Tue, 14 Jul 2009, Scott Marlowe wrote:
Are you guys doing anything that could be deemed pathological, like
full table updates on big tables over and over? Had an issue last
year where a dev left a where clause off an update to a field in one
of our biggest tables and in a few weeks the databa
On Tue, 14 Jul 2009, Scott Marlowe wrote:
Just wondering, which pgsql version, and also, do you have
autovacuum turned on?
Dang, I should have said in my initial message. 8.3.6, and autovacuum
is turned on and has plenty of log activity.
--
Sent via pgsql-performance mailing list (pgsql-perfo
Ouch hurts my eyes :) Can you see something like table_len,
dead_tuple_percent, free_percent order by dead_tuple_percent desc
limit 10 or something like that maybe?
Sorry about the pain. Didn't know what you needed to see.
Ordering by dead_tuple_percent:
db.production=> select table_name, ta
On Wed, Jul 15, 2009 at 4:03 PM, Wayne Conrad wrote:
>>> On Tue, 14 Jul 2009, Scott Marlowe wrote:
>>
>> Are you guys doing anything that could be deemed pathological, like
>> full table updates on big tables over and over? Had an issue last
>> year where a dev left a where clause off an update to
On 7/14/09 9:53 PM, "David Wilson" wrote:
> On Mon, Jul 13, 2009 at 3:31 PM, Wayne Conrad wrote:
>> Howdy. Some months back, when advised on one of these lists that it
>> should not be necessary to issue VACUUM FULL/REINDEX DATABASE, we quit
>> this nightly "maintenance" practice. We've been v
On Wed, Jul 15, 2009 at 6:51 AM, Wayne Conrad wrote:
> On Tue, 14 Jul 2009, Scott Marlowe wrote:
>>
>> Just wondering, which pgsql version, and also, do you have
>> autovacuum turned on?
>
> Dang, I should have said in my initial message. 8.3.6, and autovacuum
> is turned on and has plenty of log
On Mon, Jul 13, 2009 at 3:31 PM, Wayne Conrad wrote:
> Howdy. Some months back, when advised on one of these lists that it
> should not be necessary to issue VACUUM FULL/REINDEX DATABASE, we quit
> this nightly "maintenance" practice. We've been very happy to not
> have to do that, since it locke
On Mon, Jul 13, 2009 at 1:31 PM, Wayne Conrad wrote:
> Howdy. Some months back, when advised on one of these lists that it
> should not be necessary to issue VACUUM FULL/REINDEX DATABASE, we quit
> this nightly "maintenance" practice. We've been very happy to not
> have to do that, since it locke
