On Sat, Mar 20, 2021 at 02:12:34PM +0800, Julien Rouhaud wrote:
> On Fri, Mar 19, 2021 at 12:53:18PM -0400, Bruce Momjian wrote:
> >
> > Well, given we don't really want to support multiple query id types
> > being generated or displayed, the "error out" above should fix it.
> >
> > Let's do this --- tell extensions to error out if the query id is
> > already set, either by compute_query_id or another extension. If an
> > extension wants to generate its own query id and store is internal to
> > the extension, that is fine, but the server-displayed query id should be
> > generated once and never overwritten by an extension.
>
> Agreed, this will ensure that you won't dynamically change the queryid source.
>
> We should also document that changing it requires a restart and calling
> pg_stat_statements_reset() afterwards.
>
> v19 adds some changes, plus extra documentation for pg_stat_statements about
> the requirement for a queryid to be calculated, and a note that all documented
> details only apply for in-core source. I'm not sure if this is still the best
> place to document those details anymore though.
OK, after reading the entire thread, I don't think there are any
remaining open issues with this patch and I think this is ready for
committing. I have adjusted the doc section of the patches, attached.
I have marked myself as committer in the commitfest app and hope to
apply it in the next few days based on feedback.
--
Bruce Momjian <bruce@momjian.us> https://momjian.us
EDB https://enterprisedb.com
If only the physical world exists, free will is an illusion.