Test innodb.percona_changed_pages is unstable

Bug #1661534 reported by Laurynas Biveinis
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Percona Server moved to https://jira.percona.com/projects/PS
Status tracked in 5.7
5.5
Invalid
Undecided
Unassigned
5.6
In Progress
Low
Unassigned
5.7
In Progress
Low
Unassigned

Bug Description

On 5.6 trunk:

innodb.percona_changed_pages w3 [ fail ]
        Test ended at 2017-01-31 14:40:03

CURRENT_TEST: innodb.percona_changed_pages
--- /mnt/workspace/percona-server-5.6-trunk/BUILD_TYPE/debug/Host/centos6-64/mysql-test/suite/innodb/r/percona_changed_pages.result 2017-01-31 15:06:52.876262268 +0300
+++ /mnt/workspace/percona-server-5.6-trunk/BUILD_TYPE/debug/Host/centos6-64/build/mysql-test/var/3/log/percona_changed_pages.reject 2017-01-31 17:40:03.501594013 +0300
@@ -189,7 +189,7 @@
 SELECT COUNT(*) FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES
 WHERE START_LSN <= (@min_start_lsn - 1) OR END_LSN >= (@max_end_lsn + 1);
 COUNT(*)
-0
+1
 SELECT COUNT(*) FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES
 WHERE (START_LSN > @cond_test_max_lsn AND END_LSN < @cond_test_max_lsn)
 OR (START_LSN > @cond_test_max_lsn + 10 AND END_LSN < @cond_test_max_lsn - 10);
@@ -197,7 +197,7 @@
 0
 SELECT @total_pages=@r1_r4_pages AS should_be_1;
 should_be_1
-1
+0
 SELECT @r1_r2_pages+@r2_r3_pages+@r3_r4_pages=@r1_r4_pages AS should_be_1;
 should_be_1
 1

Tags: bitmap ci
tags: added: bitmap ci
Revision history for this message
Laurynas Biveinis (laurynas-biveinis) wrote :

With diagnostics:

include/assert.inc [No bitmap data must exist with END_LSN > @max_end_lsn]
######## Test assertion failed: No bitmap data must exist with END_LSN > @max_end_lsn ########
Dumping debug info:
Assertion text: 'No bitmap data must exist with END_LSN > @max_end_lsn'
Assertion condition: 'COUNT(*) = 0 FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn'
Assertion condition, interpolated: 'COUNT(*) = 0 FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn'
Assertion result: '0'
Assertion debug statement:
SELECT * FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn;;
space_id page_id start_lsn end_lsn
0 5 1637007 1637017
safe_process[5896]: Child process: 5897, exit: 1

Thus, it is the transaction system header that gets updated

Revision history for this message
Laurynas Biveinis (laurynas-biveinis) wrote :

Extra diagnostics were added that Innodb_max_trx_id stays the same and they did not fire, indicating that the transaction system header rewrite was not caused by its bump.

include/assert.inc [Maximum InnoDB transaction ID should not have moved during RO workload]
include/assert.inc [No bitmap data must exist with START_LSN > @max_end_lsn]
include/assert.inc [No bitmap data must exist with END_LSN > @max_end_lsn]
######## Test assertion failed: No bitmap data must exist with END_LSN > @max_end_lsn ########
Dumping debug info:
Assertion text: 'No bitmap data must exist with END_LSN > @max_end_lsn'
Assertion condition: 'COUNT(*) = 0 FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn'
Assertion condition, interpolated: 'COUNT(*) = 0 FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn'
Assertion result: '0'
Assertion debug statement:
SELECT * FROM INFORMATION_SCHEMA.INNODB_CHANGED_PAGES WHERE END_LSN > @max_end_lsn;;
space_id page_id start_lsn end_lsn
0 5 1637007 1637017
safe_process[22528]: Child process: 22529, exit: 1

Revision history for this message
Shahriyar Rzayev (rzayev-sehriyar) wrote :

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-2231

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.