| From: | Ryan Lambert <ryan(at)rustprooflabs(dot)com> |
|---|---|
| To: | Konstantin Knizhnik <k(dot)knizhnik(at)postgrespro(dot)ru> |
| Cc: | DEV_OPS <devops(at)ww-it(dot)cn>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Built-in connection pooler |
| Date: | 2019-08-08 02:48:59 |
| Message-ID: | CAN-V+g852qfYNhQyd-s+Ar0dkOLK9pzVUdmHwO-LwDdQLn0CyQ@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> I attached to this mail patch which is fixing both problems: correctly
> reports error to the client and calculates number of idle clients).
>
Yes, this works much better with max_sessions=1000. Now it's handling the
300 connections on the small server. n_idle_clients now looks accurate
with the rest of the stats here.
postgres=# SELECT n_clients, n_backends, n_idle_backends, n_idle_clients
FROM pg_pooler_state();
n_clients | n_backends | n_idle_backends | n_idle_clients
-----------+------------+-----------------+----------------
150 | 10 | 9 | 149
150 | 10 | 6 | 146
Ryan Lambert
>
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kyotaro Horiguchi | 2019-08-08 03:15:38 | Re: Small patch to fix build on Windows |
| Previous Message | Craig Ringer | 2019-08-08 02:40:29 | Re: Global temporary tables |