public inbox for [email protected]help / color / mirror / Atom feed
doubt about backend max_connections 3+ messages / 2 participants [nested] [flat]
* doubt about backend max_connections @ 2026-02-09 09:52 Luca Ferrari <[email protected]> 2026-02-12 01:17 ` Re: doubt about backend max_connections Koshino Taiki <[email protected]> 0 siblings, 1 reply; 3+ messages in thread From: Luca Ferrari @ 2026-02-09 09:52 UTC (permalink / raw) To: [email protected] Hi all, I've a doubt about configuring max_pool and num_init_children and, subsequently, PostgreSQL max_connections. In <https://www.pgpool.net/docs/46/en/html/runtime-config-connection-pooling.html; there is this sentence "Please be aware that the number of connections from Pgpool-II processes to the backends may reach num_init_children * max_pool in total. " regarding max_pool. Now, my doubt is: does this apply to all the backends (i.e., every backend will receive a max_pool * num_init_children connections or dows it apply "cluster wise", so every node will receive num_init_childre * max_pool * num_of_backends_in_pgpool? Seems to me the former, but I want to be sure. Thansk, Luca ^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: doubt about backend max_connections 2026-02-09 09:52 doubt about backend max_connections Luca Ferrari <[email protected]> @ 2026-02-12 01:17 ` Koshino Taiki <[email protected]> 2026-02-13 11:07 ` Re: doubt about backend max_connections Luca Ferrari <[email protected]> 0 siblings, 1 reply; 3+ messages in thread From: Koshino Taiki @ 2026-02-12 01:17 UTC (permalink / raw) To: Luca Ferrari <[email protected]>; +Cc: [email protected] <[email protected]> Hi, The value num_init_children × max_pool applies per backend, so every backend will receive max_pool × num_init_children connections from Pgpool-II. Please make sure that the number of connections from Pgpool-II does not exceed the maximum number of connections that each backend node can accept. Therefore, these values should satisfy one of the following formulas: When query cancellation is not considered: max_pool × num_init_children ≤ (max_connections − superuser_reserved_connections) When query cancellation is considered: max_pool × num_init_children × 2 ≤ (max_connections − superuser_reserved_connections) Because Pgpool-II opens additional connections to send cancel requests, the second formula is safer when query cancellation may occur. Taiki Koshino<[email protected]> SRA OSS K.K. TEL: 03-5979-2701 FAX: 03-5979-2702 URL: https://www.sraoss.co.jp/ ________________________________ 差出人: Luca Ferrari <[email protected]> 送信日時: 2026年2月9日 18:52 宛先: [email protected] <[email protected]> 件名: doubt about backend max_connections Hi all, I've a doubt about configuring max_pool and num_init_children and, subsequently, PostgreSQL max_connections. In <https://www.pgpool.net/docs/46/en/html/runtime-config-connection-pooling.html; there is this sentence "Please be aware that the number of connections from Pgpool-II processes to the backends may reach num_init_children * max_pool in total. " regarding max_pool. Now, my doubt is: does this apply to all the backends (i.e., every backend will receive a max_pool * num_init_children connections or dows it apply "cluster wise", so every node will receive num_init_childre * max_pool * num_of_backends_in_pgpool? Seems to me the former, but I want to be sure. Thansk, Luca ^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: doubt about backend max_connections 2026-02-09 09:52 doubt about backend max_connections Luca Ferrari <[email protected]> 2026-02-12 01:17 ` Re: doubt about backend max_connections Koshino Taiki <[email protected]> @ 2026-02-13 11:07 ` Luca Ferrari <[email protected]> 0 siblings, 0 replies; 3+ messages in thread From: Luca Ferrari @ 2026-02-13 11:07 UTC (permalink / raw) To: Koshino Taiki <[email protected]>; +Cc: [email protected] <[email protected]> On Thu, Feb 12, 2026 at 2:17 AM Koshino Taiki <[email protected]> wrote: > > Hi, > > The value num_init_children × max_pool applies per backend, > so every backend will receive max_pool × num_init_children connections from Pgpool-II. > Thanks, it is clear now. > max_pool × num_init_children × 2 ≤ (max_connections − superuser_reserved_connections) > Because Pgpool-II opens additional connections to send cancel requests, the second formula is safer when query cancellation may occur. > I didn't know this, thank you very much for pointing out. Luca ^ permalink raw reply [nested|flat] 3+ messages in thread
end of thread, other threads:[~2026-02-13 11:07 UTC | newest] Thread overview: 3+ messages (download: mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2026-02-09 09:52 doubt about backend max_connections Luca Ferrari <[email protected]> 2026-02-12 01:17 ` Koshino Taiki <[email protected]> 2026-02-13 11:07 ` Luca Ferrari <[email protected]>
This inbox is served by agora; see mirroring instructions for how to clone and mirror all data and code used for this inbox