Re: Minimal logical decoding on standbys

On 3/23/21 11:05 PM, Fabrízio de Royes Mello wrote:
>
> >
> > done in v13 attached.
> >
>
> All tests passed and everything looks good to me... just a final
> minor fix on regression tests:
>
> diff --git a/src/test/regress/expected/rules.out
> b/src/test/regress/expected/rules.out
> index b0e17d4e1d..961ec869a6 100644
> --- a/src/test/regress/expected/rules.out
> +++ b/src/test/regress/expected/rules.out
> @@ -1869,9 +1869,9 @@ pg_stat_database_conflicts| SELECT d.oid AS datid,
>      pg_stat_get_db_conflict_tablespace(d.oid) AS confl_tablespace,
>      pg_stat_get_db_conflict_lock(d.oid) AS confl_lock,
>      pg_stat_get_db_conflict_snapshot(d.oid) AS confl_snapshot,
> -    pg_stat_get_db_conflict_logicalslot(d.oid) AS confl_logicalslot,
>      pg_stat_get_db_conflict_bufferpin(d.oid) AS confl_bufferpin,
> -    pg_stat_get_db_conflict_startup_deadlock(d.oid) AS confl_deadlock
> +    pg_stat_get_db_conflict_startup_deadlock(d.oid) AS confl_deadlock,
> +    pg_stat_get_db_conflict_logicalslot(d.oid) AS confl_logicalslot
>     FROM pg_database d;
>  pg_stat_gssapi| SELECT s.pid,
>      s.gss_auth AS gss_authenticated,
>
> We moved "pg_stat_database_conflicts.confl_logicalslot" to the end of
> the column list but forgot to change the regression test expected result.

Thanks for pointing out, fixed in v14 attached.

Bertrand