Re: prevent immature WAL streaming - Mailing list pgsql-hackers

From Andres Freund
Subject Re: prevent immature WAL streaming
Date
Msg-id 20210930075152.wqu6q4pqkoospsyq@alap3.anarazel.de
Whole thread Raw
In response to Re: prevent immature WAL streaming  (Andrew Dunstan <andrew@dunslane.net>)
Responses Re: prevent immature WAL streaming  (Alvaro Herrera <alvherre@alvh.no-ip.org>)
List pgsql-hackers
Hi,

On 2021-09-29 17:04:48 -0400, Andrew Dunstan wrote:
> On 9/29/21 4:33 PM, Tom Lane wrote:
>     # Second test: a standby that receives WAL via archive/restore commands.
>     $node = PostgresNode->new('primary2');
>     $node->init(
>         has_archiving => 1,
>         extra         => ['--wal-segsize=1']);
> 
> 
> doesn't have "allows_streaming => 1".

FWIW, with that fixed I see the test hanging (e.g. [1]):

can't unlink
c:/cirrus/src/test/recovery/tmp_check/t_026_overwrite_contrecord_primary2_data/archives/000000010000000000000008.fail:
Permissiondenied at t/026_overwrite_contrecord.pl line 189.
 
### Stopping node "primary2" using mode immediate
# Running: pg_ctl -D c:/cirrus/src/test/recovery/tmp_check/t_026_overwrite_contrecord_primary2_data/pgdata -m immediate
stop
waiting for server to shut
down...........................................................................................................................
failed
pg_ctl: server does not shut down
Bail out!  command "pg_ctl -D c:/cirrus/src/test/recovery/tmp_check/t_026_overwrite_contrecord_primary2_data/pgdata -m
immediatestop" exited with value 1
 
Warning: unable to close filehandle GEN6 properly: Bad file descriptor during global destruction.


The hang seems to be fixed by uncommenting the $h->finish(). Was there a
reason you commented that out? THe test still fails, but at least it doesn't
hang anymore.

Greetings,

Andres Freund

[1]
https://api.cirrus-ci.com/v1/artifact/task/6204050896060416/tap/src/test/recovery/tmp_check/log/regress_log_026_overwrite_contrecord



pgsql-hackers by date:

Previous
From: Osumi, Takamichi/大墨 昂道
Date:
Subject: RE: Failed transaction statistics to measure the logical replication progress
Next
From: Kyotaro Horiguchi
Date:
Subject: Re: Logical replication keepalive flood