Re: aio: Async fsyncs for crash recovery and checkpointer - Mailing list pgsql-hackers

Hi,

On Tue, 22 Sept 2026 at 07:03, Yuhang Qiu <iamqyh@gmail.com> wrote:
>
> Thanks for the comparison. The main benefit of this patch series seems
> to be in checkpointing rather than crash recovery, so I think further
> testing should focus on that.

I agree. Crash recovery is useful but checkpoint's impact is much
greater than that IMO.


> > However, I think increasing the number of worker processes
> > should solve this problem.
>
> Increasing the worker count may help with worker availability, but it
> doesn't address contention for storage bandwidth.

Could you please elaborate more on this?


> One more point: when a worker calls pg_fsync() or pg_fdatasync(), it checks
> its own enableFsync. If the issuer has reloaded fsync=on but the worker
> still has the old off value, the worker can skip the sync and report
> success. Should the issuer's setting or the worker's determine whether
> the sync is performed?

Nice find. I think it makes sense to use the issuing process's
enableFsync. Done like that in v3-0002.


--
Regards,
Nazir Bilal Yavuz
Microsoft