Details
Description
rpl.rpl_parallel2 'stmt' w2 [ fail ]
Test ended at 2014-11-29 04:50:39
CURRENT_TEST: rpl.rpl_parallel2
--- /usr/local/mariadb-10.0.15-linux-i686/mysql-test/suite/rpl/r/rpl_parallel2.result 2014-11-28 15:20:23.000000000 +0100
+++ /usr/local/mariadb-10.0.15-linux-i686/mysql-test/suite/rpl/r/rpl_parallel2.reject 2014-11-29 04:50:38.000000000 +0100
@@ -10,7 +10,7 @@
Warnings:
Note 1592 Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT. Statement is unsafe because it uses a system function that may return a different value on the slave.
Seconds_Behind_Master should be zero here because the slave is fully caught up and idle.
-Seconds_Behind_Master = '0'
+Seconds_Behind_Master = '5'
include/stop_slave.inc
SET GLOBAL slave_parallel_threads=@old_parallel_threads;
include/start_slave.inc
mysqltest: Result content mismatch
Strangely, I can also see the failure on 5.5 tree http://buildbot.askmonty.org/buildbot/builders/kvm-deb-sid-amd64/builds/475/steps/test_4/logs/stdio , but can't see the actual test (and can't understand how it could have been there, if it's parallel replication...)
Gliffy Diagrams
Attachments
Issue Links
- relates to
-
MDEV-7069 Fix buildbot failures in main server trees
-
- Stalled
-
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Failure can be reproduced reliably by inserting a sleep:
=== modified file 'sql/rpl_rli.cc'
--- sql/rpl_rli.cc 2014-12-01 12:53:57 +0000
+++ sql/rpl_rli.cc 2014-12-02 08:10:32 +0000
@@ -1055,6 +1055,7 @@ void Relay_log_info::inc_group_relay_log
mysql_cond_broadcast(&data_cond);
if (!skip_lock)
mysql_mutex_unlock(&data_lock);
+my_sleep(100000);
DBUG_VOID_RETURN;
}
Pushed to 10.0.16
The reason we see the failure in the 5.5 build is the following. This is
Debian unstable (I think). In Debian unstable, there are now native MariaDB 10
packages. The Buildbot job does:
The Buildbot job has setup a repository with mariadb-server version 5.5, but
the native repository has a newer version 10.0, so the newer version is pulled
in and tested.
I would suggest that, since Debian sid now has 10.0 native, it does not make
much sense to build 5.5 for it.