On Fri, Jul 7, 2023 at 9:24 AM Rita wrote:
>
> My question are: is there a ratio of max connections and pool i should use in
> my pgbouncer config?
Short answer: No, there's no recommended ratio for PG max_connections
and pgbouncer pool size.
In pgbouncer, a client connection is matched (forwar
Rita wrote on 7/7/23 9:23 AM:
I have an application that does many db calls from a server farm.
I've increased my max connections on postgresql to 1000 and tuned the
server accordingly. However, I still get can't connect to postgresql
some times. I installed pgbouncer on few servers in the fa
I have an application that does many db calls from a server farm. I've
increased my max connections on postgresql to 1000 and tuned the server
accordingly. However, I still get can't connect to postgresql some times. I
installed pgbouncer on few servers in the farm. I pointed the traffic to
pgbou
On 7/7/23 08:27, Ron wrote:
Restarting is something I can do during a quiet point after the bulk of
the day's work is complete, but before the nightly backups.
For completeness there is the auto_explain module:
https://www.postgresql.org/docs/current/auto-explain.html
--
Adrian Klaver
ad
seems to me that sepgsql_schema_post_create should be doing a
selabel_lookup and using that default label if one exists instead of
computing the label ... hmm I'll have to think about it
On Fri, Jul 7, 2023 at 10:44 AM Ted Toth wrote:
>
> I see it now sepgsql_schema_post_create is doing it ... it
I see it now sepgsql_schema_post_create is doing it ... it's just not
doing what I think it should :(
On Fri, Jul 7, 2023 at 8:05 AM Ted Toth wrote:
>
> Also AFAICT the schema created for the temp table does not get deleted
> at the end of the session which also causes issues.
>
> On Fri, Jul 7,
On 7/7/23 10:13, Adrian Klaver wrote:
On 7/7/23 07:58, Ron wrote:
On 7/7/23 09:55, Adrian Klaver wrote:
On 7/7/23 07:51, Adrian Klaver wrote:
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
Even sett
On 7/7/23 07:58, Ron wrote:
On 7/7/23 09:55, Adrian Klaver wrote:
On 7/7/23 07:51, Adrian Klaver wrote:
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
Even setting taqs to 10KB isn't adequate, so I w
On 7/7/23 09:55, Adrian Klaver wrote:
On 7/7/23 07:51, Adrian Klaver wrote:
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
Even setting taqs to 10KB isn't adequate, so I want to significantly
bump it
On 7/7/23 09:51, Adrian Klaver wrote:
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
You can't run EXPLAIN(ANALYZE BUFFERS) if you don't have a query to run.
That's what track_activity_query_size is
On 7/7/23 07:51, Adrian Klaver wrote:
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
Even setting taqs to 10KB isn't adequate, so I want to significantly
bump it, but am
It is before coffee here, so
On 7/7/23 07:42, Ron wrote:
We've got some Very Large Queries that take a long time.
An EXPLAIN(ANALYZE BUFFERS) would go a long way here.
Even setting
taqs to 10KB isn't adequate, so I want to significantly bump it, but am
It is before coffee here, so you will need to spell out what taqs
We've got some Very Large Queries that take a long time. Even setting taqs
to 10KB isn't adequate, so I want to significantly bump it, but am concerned
about side effects of setting it to 48KB or even 64KB.
--
Born in Arizona, moved to Babylonia.
Also AFAICT the schema created for the temp table does not get deleted
at the end of the session which also causes issues.
On Fri, Jul 7, 2023 at 7:43 AM Ted Toth wrote:
>
> I'm seeing security labels set on db objects when a temporary table is
> created but they don't match the labels I've confi
I'm seeing security labels set on db objects when a temporary table is
created but they don't match the labels I've configured in my
sepgsql_contexts file. Our code doesn't actually execute a "SECURITY
LABEL ..." command for the temp tables but something else must be
doing so because I see pg_temp_
Adrian come on 🤭
This is a reduced example.
The real usecase involves many tables with the bitemporal record
However I solved using a record type ...
Il ven 7 lug 2023, 01:20 Adrian Klaver ha
scritto:
> On 7/6/23 14:52, Lorusso Domenico wrote:
> > Hello guys,
> > In my db (version 15) I've d
16 matches
Mail list logo
