rpl.rpl_slave_load_remove_tmpfile fails

Bug #1010049 reported by David Smid
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
5.1
Won't Fix
Undecided
Unassigned
5.5
Incomplete
Medium
Unassigned
5.6
Incomplete
Medium
Unassigned
5.7
Incomplete
Medium
Unassigned

Bug Description

rpl.rpl_slave_load_in 'stmt' [ pass ] 722
rpl.rpl_slave_load_remove_tmpfile 'stmt' [ fail ]
        Test ended at 2012-06-07 15:25:51

CURRENT_TEST: rpl.rpl_slave_load_remove_tmpfile
mysqltest: In included file "./include/wait_for_slave_sql_error.inc":
included from ./include/wait_for_slave_sql_error.inc at line 58:
At line 57: Slave stopped with wrong error code

The result from queries just before the failure was:
< snip >
SHOW BINLOG EVENTS IN 'master-bin.000001';
Log_name Pos Event_type Server_id End_log_pos Info
master-bin.000001 4 Format_desc 1 107 Server ver: 5.5.24-debug-log, Binlog ver: 4
master-bin.000001 107 Query 1 253 use `test`; create table t1(a int not null auto_increment, b int, primary key(a)) engine=innodb
master-bin.000001 253 Query 1 321 BEGIN
master-bin.000001 321 Intvar 1 349 INSERT_ID=1
master-bin.000001 349 Query 1 440 use `test`; insert into t1(b) values (1)
master-bin.000001 440 Intvar 1 468 INSERT_ID=2
master-bin.000001 468 Query 1 559 use `test`; insert into t1(b) values (2)
master-bin.000001 559 Intvar 1 587 INSERT_ID=3
master-bin.000001 587 Begin_load_query 1 622 ;file_id=7;block_len=12
master-bin.000001 622 Intvar 1 650 INSERT_ID=3
master-bin.000001 650 Execute_load_query 1 885 use `test`; LOAD DATA INFILE '../../std_data/rpl_loaddata.dat' INTO TABLE `t1` FIELDS TERMINATED BY '\t' ENCLOSED BY '' ESCAPED BY '\\' LINES TERMINATED BY '\n' (`a`, `b`) ;file_id=7
master-bin.000001 885 Xid 1 912 COMMIT /* xid=923 */

**** SHOW RELAYLOG EVENTS on server_1 ****
relaylog_name = 'No such row'
SHOW RELAYLOG EVENTS IN 'No such row';
Log_name Pos Event_type Server_id End_log_pos Info
**** Slave stopped with wrong error code: 1290 (expected 29, 13) ****

Revision history for this message
David Smid (datasmid) wrote :

  cmake ../Percona-Server-$latest -DBUILD_CONFIG=mysql_release -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL_PREFIX:PATH=/opt/xyz/percona-$version -DINSTALL_LAYOUT:STRING=STANDALONE -DMYSQL_DATADIR:PATH=/home/mysql/percona-5.5/data -DWITH_READLINE:BOOL=ON -DEXTRA_CHARSETS:STRING=all -DWITH_EMBEDDED_SERVER=OFF

Revision history for this message
Alexey Kopytov (akopytov) wrote :

The symptoms look identical to upstream bugs http://bugs.mysql.com/bug.php?id=60044 and http://bugs.mysql.com/bug.php?id=51190. Is the 'var' subdirectory of 'mysql-test' a symlink?

Revision history for this message
Alexey Kopytov (akopytov) wrote :

Please also provide a full command used to start the test.

Revision history for this message
David Smid (datasmid) wrote :

tests-ignored.txt:

pl.rpl_ip_mix : ipv6 not supported
rpl.rpl_ip_mix2 : ipv6 not supported
rpl.rpl_ipv4_as_ipv6 : ipv6 not supported
rpl.rpl_ipv6 : ipv6 not supported
main.ipv4_as_ipv6 : ipv6 not supported
main.ipv6 : ipv6 not supported
main.file_contents: Bug 1005787 revision-id not registered
innodb.innodb_prefix_index_liftedlimit : not suitable for build agent?

buildscript.sh:

#!/bin/bash
set -o errexit
sudo yum -y install gcc-c++ cmake bison rpm-build libaio-devel ncurses-devel gperf
patch=24
version="5.5.$patch"
perconaversion='26.0'
latest="$version-rel$perconaversion"

# http://www.percona.com/downloads/Percona-Server-5.5/Percona-Server-5.5.24-26.0/source/Percona-Server-5.5.24-rel26.0.tar.gz
wget http://www.percona.com/downloads/Percona-Server-5.5/Percona-Server-$version-$perconaversion/source/Percona-Server-$latest.tar.gz

tar -zxf Percona-Server-$latest.tar.gz

sed -i "s/%%PATCH%%/$patch/" our-percona.spec
sed -i "s/%%VERSION%%/$version/" my.cnf
sed -i "s/%%VERSION%%/$version/" our-percona.spec

rm -rf debug release tmp
mkdir -p rpm/{BUILD,RPMS,SOURCES,SPECS,SRPMS} tmp
# Build debug binaries
mkdir debug
(
  cd debug
  cmake ../Percona-Server-$latest -DBUILD_CONFIG=mysql_release -DCMAKE_BUILD_TYPE=Debug -DCMAKE_INSTALL_PREFIX:PATH=/opt/our/percona-$version -DINSTALL_LAYOUT:STRING=STANDALONE -DMYSQL_DATADIR:PATH=/home/mysql/percona-5.5/data -DWITH_READLINE:BOOL=ON -DEXTRA_CHARSETS:STRING=all -DWITH_EMBEDDED_SERVER=OFF

  make VERBOSE=1
  make test
)
(
cp tests-ignored.txt debug/mysql-test/
cp tests-ignored.txt Percona-Server-$latest/mysql-test/
cd debug/mysql-test/
./mysql-test-run.pl --skip-test-list=tests-ignored.txt
)
# Build release binaries and create final package
mkdir release
(
  cd release
  cmake ../Percona-Server-$latest -DBUILD_CONFIG=mysql_release -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX:PATH=/opt/our/percona-$version -DINSTALL_LAYOUT:STRING=STANDALONE -DMYSQL_DATADIR:PATH=/home/mysql/percona-5.5/data -DWITH_READLINE:BOOL=ON -DEXTRA_CHARSETS:STRING=all -DENABLE_DEBUG_SYNC:BOOL=OFF -DWITH_EMBEDDED_SERVER=OFF
  make VERBOSE=1 package
)

cp release/mysql-$version-linux-x86_64.tar.gz tmp
(
  cd tmp
  tar xfz mysql-$version-linux-x86_64.tar.gz
  mv mysql-$version-linux-x86_64 our-percona-$version
  cp ../my.cnf our-percona-$version/support-files
  tar cfz percona-$patch-linux-x86_64.tgz our-percona-$version
)
mv tmp/percona-$patch-linux-x86_64.tgz rpm/SOURCES
cp our-percona.spec rpm/SPECS
rpmbuild -v --define="_topdir $PWD/rpm" --define="_tmppath $PWD/tmp" -bb rpm/SPECS/our-percona.spec
cp rpm/RPMS/x86_64/our-percona-*.rpm .

Revision history for this message
David Smid (datasmid) wrote :

This form wraps!

Revision history for this message
Valerii Kravchuk (valerii-kravchuk) wrote :

I do not see failure of this tests in Jenkins for recent builds, so two questions remain:

1. Is the 'var' subdirectory of 'mysql-test' a symlink?
2. Is this still repeatable for you with recent versions, 5.5.30+?

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-1257

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.