Re: add checkpoint stats of snapshot and mapping files of pg_logical dir

On Wed, Mar 16, 2022 at 1:47 AM Nathan Bossart <nathandbossart(at)gmail(dot)com> wrote:
>
> On Tue, Mar 15, 2022 at 11:04:26AM +0900, Michael Paquier wrote:
> > On Mon, Mar 14, 2022 at 03:54:19PM +0530, Bharath Rupireddy wrote:
> >> At times, the snapshot or mapping files can be large in number and one
> >> some platforms it takes time for checkpoint to process all of them.
> >> Having the stats about them in server logs can help us better analyze
> >> why checkpoint took a long time and provide a better RCA.
> >
> > Do you have any numbers to share regarding that? Seeing information
> > about 1k WAL segments being recycled and/or removed by a checkpoint
> > where the operation takes dozens of seconds to complete because we can
> > talk about hundred of gigs worth of files moved around. If we are
> > talking about 100~200 files up to 10~20kB each for snapshot and
> > mapping files, the information has less value, worth only a portion of
> > one WAL segment.
>
> I don't have specific numbers to share, but as noted elsewhere [0], I
> routinely see lengthy checkpoints that spend a lot of time in these cleanup
> tasks.
>
> [0] https://postgr.es/m/18ED8B1F-7F5B-4ABF-848D-45916C938BC7%40amazon.com

I too don't have numbers in hand to share. But I have come across
cases where checkpoints spend a good amount of time cleaning
up/processing snapshots or mapping files.

Regards,
Bharath Rupireddy.