public inbox for [email protected]help / color / mirror / Atom feed
TPS in direct Postgres Vs TPS in pgbouncer 3+ messages / 3 participants [nested] [flat]
* TPS in direct Postgres Vs TPS in pgbouncer @ 2026-06-21 13:24 Raj <[email protected]> 2026-06-21 14:37 ` RE: TPS in direct Postgres Vs TPS in pgbouncer Jorge Batista <[email protected]> 2026-06-21 19:19 ` Re: TPS in direct Postgres Vs TPS in pgbouncer Jeffery Janes <[email protected]> 0 siblings, 2 replies; 3+ messages in thread From: Raj @ 2026-06-21 13:24 UTC (permalink / raw) To: Pgsql-admin <[email protected]> Hi All, When the application team is using posytgres directly for 17.9 version, they get necessary TPS. When they use pgbouncer, they get 50% reduction in TPS. What are general advices? My set up is 3 node PATRONI, where node3 is DR and manual failover. Two nodes (node1 and node2) are in site A ..having haproxy, pgnouncer is bound with haproxy. ^ permalink raw reply [nested|flat] 3+ messages in thread
* RE: TPS in direct Postgres Vs TPS in pgbouncer 2026-06-21 13:24 TPS in direct Postgres Vs TPS in pgbouncer Raj <[email protected]> @ 2026-06-21 14:37 ` Jorge Batista <[email protected]> 1 sibling, 0 replies; 3+ messages in thread From: Jorge Batista @ 2026-06-21 14:37 UTC (permalink / raw) To: Raj <[email protected]>; Pgsql-admin <[email protected]> Raj, when you turned on pgbouncer, did you check if it's set to "transaction mode" & not "session mode"? Also worth checking if pgbouncer itself is maxing out a CPU core during the slow test, both are common causes of a TPS drop like that, separate from Postgres itself. Regards, jorge batista From: Raj <[email protected]> Sent: Sunday, June 21, 2026 9:25 AM To: Pgsql-admin <[email protected]> Subject: TPS in direct Postgres Vs TPS in pgbouncer EXTERNAL: This email originated from outside of our organization. Do not click links or open attachments unless you recognize the sender and know the content is safe. Hi All, When the application team is using posytgres directly for 17.9 version, they get necessary TPS. When they use pgbouncer, they get 50% reduction in TPS. What are general advices? My set up is 3 node PATRONI, where node3 is DR and manual failover. Two nodes (node1 and node2) are in site A ..having haproxy, pgnouncer is bound with haproxy. ________________________________ Notice: This e-mail is intended only for the named person or entity to which it is addressed and contains valuable business information that is privileged, confidential and/or otherwise protected from disclosure. Dissemination, distribution or copying of this e-mail or the information herein by anyone other than the intended recipient, or an employee, or agent responsible for delivering the message to the intended recipient, is strictly prohibited. All contents are the copyright property of the sender. If you are not the intended recipient, you are nevertheless bound to respect the sender's worldwide legal rights. We require that unintended recipients delete the e-mail and destroy all electronic copies in their system, retaining no copies in any media. If you have received this e-mail in error, please immediately notify us by returning the e-mail to the sender. We appreciate your cooperation. ^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: TPS in direct Postgres Vs TPS in pgbouncer 2026-06-21 13:24 TPS in direct Postgres Vs TPS in pgbouncer Raj <[email protected]> @ 2026-06-21 19:19 ` Jeffery Janes <[email protected]> 1 sibling, 0 replies; 3+ messages in thread From: Jeffery Janes @ 2026-06-21 19:19 UTC (permalink / raw) To: Raj <[email protected]>; +Cc: Pgsql-admin <[email protected]> On Sun, Jun 21, 2026 at 9:24 AM Raj <[email protected]> wrote: > Hi All, > > When the application team is using posytgres directly for 17.9 version, > they get necessary TPS. > > When they use pgbouncer, they get 50% reduction in TPS. > > What are general advices? > What is the transaction being used? If it is something very simple, like pgbench's built-in -S, then I think this is expected. Injecting pgbouncer doubles the amount of network/IPC thrashing needed. If you have a meaty transaction, this might not matter but on a very light transaction you might be doubling something which is already counting for 90% of the time. Cheers, Jeff ^ permalink raw reply [nested|flat] 3+ messages in thread
end of thread, other threads:[~2026-06-21 19:19 UTC | newest] Thread overview: 3+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2026-06-21 13:24 TPS in direct Postgres Vs TPS in pgbouncer Raj <[email protected]> 2026-06-21 14:37 ` Jorge Batista <[email protected]> 2026-06-21 19:19 ` Jeffery Janes <[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