Re: Resetting spilled txn statistics in pg_stat_replication

On Wed, Sep 9, 2020 at 3:20 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>
> On Tue, Sep 8, 2020 at 7:02 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
> >
> > Comments on the latest patch:
> > =============================
> >
>
> Apart from the comments I gave yesterday, another thing I was
> wondering is how to write some tests for this patch. The two ideas I
> could think of are as follows:
>
> 1. One idea was to keep these stats for each WALSender as it was in
> the commit that we reverted as b074813d48. If we had that then we can
> query the stats for tests added in commit 58b5ae9d62. I am not sure
> whether we want to display it in view pg_stat_replication but it would
> be a really good way to test the streamed and serialized transactions
> in a predictable manner.
>
> 2. Then the second way is to try doing something similar to what we do
> in src/test/regress/sql/stats.sql
>
> I think we should do both if possible.
>

I have made a few comment changes on top of your last version. If you
are fine with these then include them with the next version of your
patch.

--
With Regards,
Amit Kapila.