Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commit41ffd9d

Browse files
committed
Test restartpoints in archive recovery.
v14 commit1f95181 and its v13equivalent caused timing-dependent failures in archive recovery, atrestartpoints. The symptom was "invalid magic number 0000 in logsegment X, offset 0", "unexpected pageaddr X in log segment Y, offset 0"[X < Y], or an assertion failure. Commit3635a0a and predecessors back-patchedv15 changes to fix that. This test reproduces the problemprobabilistically, typically in less than 1000 iterations of the test.Hence, buildfarm and CI runs would have surfaced enough failures to getattention within a day.Reported-by: Arun Thirupathi <arunth@google.com>Discussion:https://postgr.es/m/20250306193013.36.nmisch@google.comBackpatch-through: 13
1 parenta61b67c commit41ffd9d

File tree

1 file changed

+57
-0
lines changed

1 file changed

+57
-0
lines changed
Lines changed: 57 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,57 @@
1+
2+
# Copyright (c) 2024-2025, PostgreSQL Global Development Group
3+
4+
# Test restartpoints during archive recovery.
5+
use strict;
6+
use warnings;
7+
8+
use PostgreSQL::Test::Cluster;
9+
use PostgreSQL::Test::Utils;
10+
use Test::More;
11+
12+
my$archive_max_mb = 320;
13+
my$wal_segsize = 1;
14+
15+
# Initialize primary node
16+
my$node_primary = PostgreSQL::Test::Cluster->new('primary');
17+
$node_primary->init(
18+
has_archiving=> 1,
19+
allows_streaming=> 1,
20+
extra=> ['--wal-segsize'=>$wal_segsize ]);
21+
$node_primary->start;
22+
my$backup_name ='my_backup';
23+
$node_primary->backup($backup_name);
24+
25+
$node_primary->safe_psql('postgres',
26+
('DO $$BEGIN FOR i IN 1..' .$archive_max_mb /$wal_segsize)
27+
.' LOOP CHECKPOINT; PERFORM pg_switch_wal(); END LOOP; END$$;');
28+
29+
# Force archiving of WAL file containing recovery target
30+
my$until_lsn =$node_primary->lsn('write');
31+
$node_primary->safe_psql('postgres',"SELECT pg_switch_wal()");
32+
$node_primary->stop;
33+
34+
# Archive recovery
35+
my$node_restore = PostgreSQL::Test::Cluster->new('restore');
36+
$node_restore->init_from_backup($node_primary,$backup_name,
37+
has_restoring=> 1);
38+
$node_restore->append_conf('postgresql.conf',
39+
"recovery_target_lsn = '$until_lsn'");
40+
$node_restore->append_conf('postgresql.conf',
41+
'recovery_target_action = pause');
42+
$node_restore->append_conf('postgresql.conf',
43+
'max_wal_size =' . 2 *$wal_segsize);
44+
$node_restore->append_conf('postgresql.conf','log_checkpoints = on');
45+
46+
$node_restore->start;
47+
48+
# Wait until restore has replayed enough data
49+
my$caughtup_query =
50+
"SELECT '$until_lsn'::pg_lsn <= pg_last_wal_replay_lsn()";
51+
$node_restore->poll_query_until('postgres',$caughtup_query)
52+
ordie"Timed out while waiting for restore to catch up";
53+
54+
$node_restore->stop;
55+
ok(1,'restore caught up');
56+
57+
done_testing();

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp