PG17 failing tests (DST related?)

Started by Dave Pageover 1 year ago3 messageshackers
Jump to latest
#1Dave Page
dpage@pgadmin.org

For about a week, the Windows builds I've been running in Github have been
failing for PG17:
https://github.com/dpage/winpgbuild/actions/workflows/postgresql.yml. The
consistently failing tests are:

postgresql:recovery / recovery/027_stream_regress
postgresql:pg_upgrade / pg_upgrade/002_pg_upgrade
postgresql:regress / regress/regress

which all appear to have the same failure in jsonb_jsonpath:

=====
diff --strip-trailing-cr -U3
D:/a/winpgbuild/winpgbuild/postgresql-17.0/src/test/regress/expected/jsonb_jsonpath.out
D:/a/winpgbuild/winpgbuild/postgresql-17.0/build/testrun/regress/regress/results/jsonb_jsonpath.out
---
D:/a/winpgbuild/winpgbuild/postgresql-17.0/src/test/regress/expected/jsonb_jsonpath.out
2024-09-23 20:02:53.000000000 +0000
+++
D:/a/winpgbuild/winpgbuild/postgresql-17.0/build/testrun/regress/regress/results/jsonb_jsonpath.out
2024-11-11 04:16:42.243042300 +0000
@@ -2637,7 +2637,7 @@
select jsonb_path_query_tz('"12:34:56"', '$.time_tz().string()');
jsonb_path_query_tz
---------------------
- "12:34:56-07:00"
+ "12:34:56-08:00"
(1 row)

select jsonb_path_query('"12:34:56"', '$.time().string()');
=====

It seems likely this is related to the DST change in the US, as the issue
started occurring on 4th November.

--
Dave Page
pgAdmin: https://www.pgadmin.org
PostgreSQL: https://www.postgresql.org

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: Dave Page (#1)
Re: PG17 failing tests (DST related?)

Dave Page <dpage@pgadmin.org> writes:

For about a week, the Windows builds I've been running in Github have been
failing for PG17:

Should be fixed as of a35801915 ?

regards, tom lane

#3Dave Page
dpage@pgadmin.org
In reply to: Tom Lane (#2)
Re: PG17 failing tests (DST related?)

On Mon, 11 Nov 2024 at 15:35, Tom Lane <tgl@sss.pgh.pa.us> wrote:

Dave Page <dpage@pgadmin.org> writes:

For about a week, the Windows builds I've been running in Github have

been

failing for PG17:

Should be fixed as of a35801915 ?

Ah, yes - it does appear to be fine with head. I guess I'll have to put up
with the error messages until 17.1!

Thanks.

--
Dave Page
pgAdmin: https://www.pgadmin.org
PostgreSQL: https://www.postgresql.org