Re: Race condition in recovery?
Andrew Dunstan <andrew@dunslane.net>
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: 2021-06-14T19:46:57Z
Lists: pgsql-hackers
On 6/14/21 3:32 PM, Andrew Dunstan wrote:
> On 6/14/21 1:50 PM, Andrew Dunstan wrote:
>> 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.
>
>
> Actually, I take it back, it does work for a file. I'll change it. I
> probably did this when something else wasn't working.
So, will you feel happier with this applied? I haven't tested it yet but
I'm confident it will work.
diff --git a/src/test/recovery/t/025_stuck_on_old_timeline.pl b/src/test/recovery/t/025_stuck_on_old_timeline.pl
index e4e58cb8ab..3e19bc4c50 100644
--- a/src/test/recovery/t/025_stuck_on_old_timeline.pl
+++ b/src/test/recovery/t/025_stuck_on_old_timeline.pl
@@ -24,11 +24,11 @@ my $node_primary = get_new_node('primary');
# the timeline history file reaches the archive but before any of the WAL files
# get there.
$node_primary->init(allows_streaming => 1, has_archiving => 1);
-my $perlbin = $^X;
-if ($^O eq 'msys')
-{
- $perlbin = TestLib::perl2host(dirname($^X)) . '\\' . basename($^X);
-}
+
+# Note: consistent use of forward slashes here avoids any escaping problems
+# that arise from use of backslashes. That means we need to double-quote all
+# the paths in the archive_command
+my $perlbin = TestLib::perl2host(^X);
$perlbin =~ s!\\!/!g if $TestLib::windows_os;
my $archivedir_primary = $node_primary->archive_dir;
$archivedir_primary =~ s!\\!/!g if $TestLib::windows_os;
@@ -36,6 +36,8 @@ $node_primary->append_conf('postgresql.conf', qq(
archive_command = '"$perlbin" "$FindBin::RealBin/cp_history_files" "%p" "$archivedir_primary/%f"'
wal_keep_size=128MB
));
+# make sure that Msys perl doesn't complain about difficulty in setting locale
+# when called this way.
local $ENV{PERL_BADLANG}=0;
$node_primary->start;
cheers
andrew
--
Andrew Dunstan
EDB: https://www.enterprisedb.com
Commits
-
Further refinement of stuck_on_old_timeline recovery test
- e46f6a078c1a 9.6.23 landed
- 1dc53c914fe8 10.18 landed
- c0a7587807dd 11.13 landed
- fb3d6b0e1aa4 12.8 landed
- d906d106f854 13.4 landed
- 54a5ed220169 14.0 landed
-
Adjust new test case to set wal_keep_segments.
- 6cd0e55f9ffb 9.6.23 landed
- 1c2c6a9a299c 10.18 landed
- 534b9be805b4 11.13 landed
- 2208d71a0099 12.8 landed
-
Adjust new test case to set wal_keep_size.
- 3465328aa19e 13.4 landed
- 4dcb1d087aeb 14.0 landed
-
Fix corner case failure of new standby to follow new primary.
- 6eb5b9ae3917 9.6.23 landed
- 31bda6a22ad1 10.18 landed
- ca158c168ea3 11.13 landed
- 5d950c721d38 12.8 landed
- 082656429215 13.4 landed
- caba8f0d43fb 14.0 landed
-
Back-port a few PostgresNode.pm methods.
- 75212a854f99 9.6.23 landed
-
Allow PostgresNode.pm's backup method to accept backup_options.
- 9778a191a4c0 10.18 landed
- 38982b8b7bd9 11.13 landed
- 6fc2febc28ae 12.8 landed
- 99a0a2ada8c5 13.4 landed
-
Add end-to-end testing of pg_basebackup's tar-format output.
- 081876d75ea1 14.0 cited
-
Allow PostgresNode.pm tests to wait for catchup
- fb093e4cb36f 10.0 cited
-
Delay reading timeline history file until it's fetched from master.
- ee994272ca50 9.3.0 cited