From: Manish Kumar Date: August 10 2012 10:21am Subject: bzr push into mysql-5.6 branch (manish.4.kumar:4129 to 4130) Bug#13886150 List-Archive: http://lists.mysql.com/commits/144540 X-Bug: 13886150 Message-Id: <201208101023.q7AANETA005908@acsmt357.oracle.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit 4130 Manish Kumar 2012-08-10 BUG#13886150 - RPL.RPL_INCIDENT FAILS ON PB2 SPORADICALLY DUE TO TIMEOUT Problem: rpl.rpl_incident fails on pb2 sporadically due to timeout in include/wait_for_slave_param.inc. Apart from generic linux(linux2.6-x86-64bit, where we have knows timeout issues), the failure was observed on gcov-linux-x86_64 and rhel5-x86-32bit. The test is passing in retry attempts in just 257 milliseconds, it means it might be hanging sometime in include/wait_for_slave_param.inc. Fix: Adding a force-restart option to the test will restart the server before starting to execute the test. modified: mysql-test/suite/rpl/t/rpl_incident-master.opt 4129 Astha Pareek 2012-08-10 BUG#14340819 MTR NOT MASKING THE WARNING NUMBER AFTER CHANGE MASTER IN RPL_SLAVE_NET_TIMEOUT Problem: MTR does not mask the warning number being thrown after change master in rpl_slave_net_timeout.test. Ideally all such warning numbers should be masked with #. But for mysql-test/suite/large_tests/r/rpl_slave_net_timeout.test, 1st warning number is masked properly, but it fails to mask the 2nd warning number, causes a result mismatch for this test. Fix: Masked the warning number for rpl_slave_net_timeout.test modified: mysql-test/suite/large_tests/r/rpl_slave_net_timeout.result mysql-test/suite/large_tests/t/rpl_slave_net_timeout.test === modified file 'mysql-test/suite/rpl/t/rpl_incident-master.opt' --- a/mysql-test/suite/rpl/t/rpl_incident-master.opt 2007-06-27 12:28:02 +0000 +++ b/mysql-test/suite/rpl/t/rpl_incident-master.opt 2012-08-10 10:19:16 +0000 @@ -1 +1 @@ ---loose-debug=+d,incident_database_resync_on_replace +--force-restart --loose-debug=+d,incident_database_resync_on_replace No bundle (reason: useless for push emails).