Bug #47477 PB2: rpl_start_stop_slave failures
Submitted: 21 Sep 2009 10:13 Modified: 1 Feb 2010 11:24
Reporter: Luis Soares Email Updates:
Status: Can't repeat Impact on me:
None 
Category:Tests: Replication Severity:S3 (Non-critical)
Version: OS:Any
Assigned to: Serge Kozlov CPU Architecture:Any
Tags: pb-2
Triage: Triaged: D3 (Medium)

[21 Sep 2009 10:13] Luis Soares
Description:
rpl_start_stop_slave fails on PB2 with the following symptoms:

=== Symptom 1 ===

Where 	Build 	Test 	Run 	Suite 	Case 	Mode 	When
mysql-pe
532264	binary-max-linux-x86_64-tar-gz
532340 tyr32	test-max-linux-x86_64
532408 tyr36	rpl_binlog_row	rpl	rpl_start_stop_slave	None	2009-09-18 10:22:00

link: http://pb2.norway.sun.com/web.py?template=mysql_show_test_failure&test_failure_id=2327602

rpl.rpl_start_stop_slave                 w7 [ fail ]
        Test ended at 2009-09-18 11:45:34

CURRENT_TEST: rpl.rpl_start_stop_slave
--- /export/home/pb2/test/sb_6-753550-1253264352.53/mysql-6.0.14-alpha-linux-x86_64-test/mysql-test/suite/rpl/r/rpl_start_stop_slave.result	2009-09-18 11:55:00.000000000 +0300
+++ /export/home/pb2/test/sb_6-753550-1253264352.53/mysql-6.0.14-alpha-linux-x86_64-test/mysql-test/suite/rpl/r/rpl_start_stop_slave.reject	2009-09-18 12:45:34.000000000 +0300
@@ -7,6 +7,7 @@
 create table t1(n int);
 stop slave;
 start slave;
+Timeout in wait_condition.inc for SELECT COUNT(*) > 0 FROM t1
 stop slave io_thread;
 start slave io_thread;
 drop table t1;

mysqltest: Result content mismatch

=== Symptom 2 ===

Where 	Build 	Test 	Run 	Suite 	Case 	Mode 	When
mysql-6.0-codebase-bugfixing
532911	binary-max-sol10-sparc64-tar-gz
533099 takk-z4	test-max-sol10-sparc64
533398 tor04-z6	ps_row	rpl	rpl_start_stop_slave	None	2009-09-18 17:38:23

link: http://pb2.norway.sun.com/web.py?template=mysql_show_test_failure&test_failure_id=2328549

rpl.rpl_start_stop_slave                 w5 [ fail ]
        Test ended at 2009-09-18 22:45:42

CURRENT_TEST: rpl.rpl_start_stop_slave
mysqltest: In included file "./include/wait_condition.inc": At line 42: Error running query 'select count(*) = 1 from t2m': 2013 Lost connection to MySQL server during query

The result from queries just before the failure was:
< snip >
start slave;
stop slave io_thread;
start slave io_thread;
drop table t1;
create table t1i(n int primary key) engine=innodb;
create table t2m(n int primary key) engine=myisam;
begin;
insert into t1i values (1);
insert into t1i values (2);
insert into t1i values (3);
commit;
begin;
insert into t1i values (5);
begin;
insert into t1i values (4);
insert into t2m values (1);
insert into t1i values (5);
commit;
zero
0

More results from queries before failure can be found in /export/home/pb2/test/sb_0-754690-1253294635.16/mysql-6.0.14-alpha-solaris10-sparc-test/mysql-test/var-ps_row/5/log/rpl_start_stop_slave.log

How to repeat:
Symptom 1: http://pb2.norway.sun.com/web.py?template=mysql_show_test_failure&test_failure_id=2327602
Symptom 2: http://pb2.norway.sun.com/web.py?template=mysql_show_test_failure&test_failure_id=2328549

Suggested fix:
n/a
[1 Feb 2010 11:24] Serge Kozlov
The analyzing show following:
1. The original test case 'rpl.rpl_start_stop_slave' doesn't exist in 6.0 trees where failures were found.
2. It moved into extra tests and run only in statement binlog format (rpl_stm_str_stop_slave). Original failure happened under row binlog format.