Re: Intermittent failure with t/003_logical_slots.pl test on windows

Thanks for working on it. I tested the patch on my system and it
resolved the issue with commands running -V (version check).

As you mentioned, I am also still seeing intermittent errors even with
the patch as below -
in 'pg_upgrade/002_pg_upgrade' -

# Running: pg_upgrade --no-sync -d
D:\Project\pg2\postgres\build/testrun/pg_upgrade/002_pg_upgrade\data/t_002_pg_upgrade_old_node_data/pgdata
-D D:\Project\pg2\postgres\build/testrun/pg_upgrade/002_pg_upgrade\data/t_002_pg_upgrade_new_node_data/pgdata
-b D:/Project/pg2/postgres/build/tmp_install/Project/pg2/postgresql/bin
-B D:/Project/pg2/postgres/build/tmp_install/Project/pg2/postgresql/bin
-s 127.0.0.1 -p 56095 -P 56096 --copy --check
Performing Consistency Checks
-----------------------------
Checking cluster versions ok

The source cluster lacks cluster state information:
Failure, exiting
[12:37:38.666](3.317s) not ok 12 - run of pg_upgrade --check for new instance
[12:37:38.666](0.000s) # Failed test 'run of pg_upgrade --check for
new instance'
# at D:/Project/pg2/postgres/src/bin/pg_upgrade/t/002_pg_upgrade.pl line 375.

and in 'pg_upgrade/003_logical_slots' -

[12:35:33.773](0.001s) not ok 6 - run of pg_upgrade of old cluster
with slots having unconsumed WAL records stdout /(?^:Your installation
contains logical replication slots that can't be upgraded.)/
[12:35:33.773](0.000s) # Failed test 'run of pg_upgrade of old
cluster with slots having unconsumed WAL records stdout /(?^:Your
installation contains logical replication slots that can't be
upgraded.)/'
# at D:/Project/pg2/postgres/src/bin/pg_upgrade/t/003_logical_slots.pl
line 102.
[12:35:33.773](0.000s) # 'Performing Consistency Checks
# -----------------------------
# Checking cluster versions ok
#
# The target cluster lacks cluster state information:
# Failure, exiting

It seems 'Performing Consistency Checks' fail due to a lack of some
information and possible that it can also be fixed on the same lines.

--
Thanks,
Nisha