Vladimir,
pgbouncer works with pl/proxy in transaction pooling mode. A wide spread
phrase that statement mode is for plproxy does not mean any limitations for
transaction pooling mode until you have atocommit on client. Anyway, try to
reduce connections.
try to set your autovacuum a bit more aggr
13.02.2014, в 13:29, Ilya Kosmodemiansky написал(а):
> Vladimir,
>
> And, any effect on your problem?
It worked without problems longer than previous configuration but repeated
again several minutes ago :(
>
> On Thu, Feb 13, 2014 at 9:35 AM, Бородин Владимир wrote:
>> I have limited max c
On Thu, Feb 13, 2014 at 11:23 AM, Бородин Владимир wrote:
> Cool. How much ram do you have on your server and what is the size of the
> database?
It has 200 GiB of memory for 100 GB database at that time. We had migrated
the database from MySQL, that was the reason of overmuch resources. I do
not
Vladimir,
And, any effect on your problem?
On Thu, Feb 13, 2014 at 9:35 AM, Бородин Владимир wrote:
> I have limited max connections to 1000, reduced shared buffers to 8G and
> restarted postgres.
1000 is still to much in most cases. With pgbouncer in transaction
pooling mode normaly pool size
Cool. How much ram do you have on your server and what is the size of the
database?
My settings of 32 GB was set due to recommendation of 25% from here [1]. Right
now I am going to investigate the content of buffer cache with pg_buffercache
and maybe OS page cache with recently released pg_stat
2014-02-12 18:59, Бородин Владимир :
> I have read several topics [1, 2, 3, 4] with similar problems but haven't
> find a good solution. Below is some more diagnostics.
I reported the second one. The diagnostics was very similar to yours.
I think a lot people experienced this problem with big serv
I have limited max connections to 1000, reduced shared buffers to 8G and
restarted postgres.
The logs say that checkpoint finishes in 2,5 minutes (as expected due to
default checkpoint_completion_target = 0.5) with no IO spikes so I don't want
to increase checkpoint_timeout of checkpoint_segme
