Re: Parallelizing startup with many databases
| От | Jelte Fennema-Nio |
|---|---|
| Тема | Re: Parallelizing startup with many databases |
| Дата | |
| Msg-id | CAGECzQSu1BeW3EvRwUWGYL+EBHzDhLCRW3LKcqTU-Y4_44myZA@mail.gmail.com обсуждение исходный текст |
| Ответ на | Parallelizing startup with many databases (Babak Ghadiri <bbkghadiri6@gmail.com>) |
| Ответы |
Re: Parallelizing startup with many databases
|
| Список | pgsql-hackers |
On Fri, Jan 2, 2026, 08:55 Babak Ghadiri <bbkghadiri6@gmail.com> wrote:
Hello,
I hope you are doing well.
In PostgreSQL 16, startup appears to initialize databases sequentially and
primarily uses a single CPU core. In clusters with a very large number of
databases (around 5,000 in our case), this results in noticeably long
startup times after restarts or crash recovery.
You probably want to consider setting:
recovery_init_sync_method=syncfsI'm 99% certain that that will solve your problem.
PS It took me way to long to find that setting. I think we should move it from the error handling docs page to the page with all of the other recovery settings. https://www.postgresql.org/docs/current/runtime-config-wal.html#RUNTIME-CONFIG-WAL-RECOVERY
В списке pgsql-hackers по дате отправления: