Re: Windows vs recovery tests

Hi,

On 2022-01-12 18:25:26 -0800, Andres Freund wrote:
> There's one interesting patch that also times out just on windows, albeit in
> another test group:
> https://cirrus-ci.com/github/postgresql-cfbot/postgresql/commitfest/36/2096
>
> This IMO looks likely to be a bug in psql introduced by that patch.

vcregress doesn't say which tests it's about to run unfortunately, but
comparing a successful run (on another branch) says that the test running
after pgbench are the psql tests.

I pushed a branch to my github repository containing cfbot's commit and one
that runs the psql tests in isolation, under a timeout... Which predictably
failed. But at least we see the logs...

https://cirrus-ci.com/task/6723083204558848?logs=psql_test_tcp#L15

based on the log files it looks like psql's 001_basic test did run

# Test clean handling of unsupported replication command responses
psql_like(
$node,
'handling of unexpected PQresultStatus',
'START_REPLICATION 0/0',
undef, qr/unexpected PQresultStatus: 8$/);

https://api.cirrus-ci.com/v1/artifact/task/6723083204558848/log/src/bin/psql/tmp_check/log/001_basic_main.log
2022-01-13 03:28:45.973 GMT [604][walsender] [001_basic.pl][3/0:0] STATEMENT: START_REPLICATION 0/0

https://api.cirrus-ci.com/v1/artifact/task/6723083204558848/tap/src/bin/psql/tmp_check/log/regress_log_001_basic

the last log entry in tap log is

ok 23 - \help with argument: stdout matches

So it looks like psql is hanging somewhere after that. I assume with an error
popup that nobody can click on :/.

Greetings,

Andres Freund