Horton

Re: Race condition in recovery?

Andrew Dunstan <andrew@dunslane.net>

View thread View on postgr.es
From: Andrew Dunstan <andrew@dunslane.net>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Michael Paquier <michael@paquier.xyz>, Dilip Kumar <dilipbalaut@gmail.com>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, Heikki Linnakangas <hlinnaka@iki.fi>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-06-14T17:50:26Z
Lists: pgsql-hackers 
On 6/14/21 1:11 PM, Robert Haas wrote:
> On Mon, Jun 14, 2021 at 12:56 PM Andrew Dunstan <andrew@dunslane.net> wrote:
>> $^X is not at all broken.
>>
>> The explanation here is pretty simple - the argument to perl2host is
>> meant to be a directory. If we're going to accomodate plain files then
>> we have some more work to do in TestLib.
> This explanation seems to contradict the documentation in TestLib.pm,
> which makes no mention of any such restriction.


Heres a snippet:


    sub perl2host
    {
        my ($subject) = @_;
        ...
        if (chdir $subject)
     

Last time I looked you can't chdir to anything except a directory.


>
>>> +local $ENV{PERL_BADLANG}=0;
>>>
>>> Similarly here. There's not a single other reference to PERL_BADLANG
>>> in the repository, so if we need this one here, there should be a
>>> comment explaining why this is different from all the places where we
>>> don't need it.
>> Here's why this is different: this is the only place that we invoke the
>> msys perl in this way (i.e. from a non-msys aware environment - the
>> binaries we build are not msys-aware). We need to do that if for no
>> other reason than that it might well be the only perl available. Doing
>> so makes it complain loudly about missing locale info. Setting this
>> variable makes it shut up. I can add a comment on that if you like.
> Yes, please, but perhaps you'd like to post patches for discussion
> first instead of committing directly.


I was trying to get the buildfarm green again. There have been plenty of
times when small patches directly for such fixes have been committed
directly. And that's the only circumstance when I do.



cheers


andrew

Commits

  1. Further refinement of stuck_on_old_timeline recovery test

  2. Adjust new test case to set wal_keep_segments.

  3. Adjust new test case to set wal_keep_size.

  4. Fix corner case failure of new standby to follow new primary.

  5. Back-port a few PostgresNode.pm methods.

  6. Allow PostgresNode.pm's backup method to accept backup_options.

  7. Add end-to-end testing of pg_basebackup's tar-format output.

  8. Allow PostgresNode.pm tests to wait for catchup

  9. Delay reading timeline history file until it's fetched from master.