Hi.
Testing mysql-master-ha (with 3 slaves and one master), I discovered that the
new master will still be seen as a slave and masterha_manager then refuses to
start.
It also won't remove the failed master from the config when I run:
# masterha_manager --remove_dead_master_conf --conf=/etc/mha/app1.cnf
This is part of the log telling that mysql-master-ha failed to remove the slave
part from the new master and that it still runs as slave:
Tue Sep 25 14:25:45 2012 - [info] * Phase 5: New master cleanup phease..
Tue Sep 25 14:25:45 2012 - [info]
Tue Sep 25 14:25:45 2012 - [info] Resetting slave info on the new master..
Tue Sep 25 14:25:45 2012 - [error][/usr/share/perl5/vendor_perl/MHA/Server.pm,
ln674] SHOW SLAVE STATUS shows new master replicates from somewhere. Check for
details!
Tue Sep 25 14:25:45 2012 - [error][/usr/share/perl5/vendor_perl/MHA/Server.pm,
ln688] db02.db.cert.fronter.net: Resetting slave info failed.
Tue Sep 25 14:25:45 2012 -
[error][/usr/share/perl5/vendor_perl/MHA/MasterFailover.pm, ln1537] Master
failover to db02.mynetwork.net(11.22.33.2:3306) done, but recovery on slave
partially failed.
Tue Sep 25 14:25:45 2012 - [info]
This is output of show slave status:
mysql> show slave status\G;
*************************** 1. row ***************************
Slave_IO_State:
Master_Host: db01.mynetwork.net
Master_User: replica
Master_Port: 3306
Connect_Retry: 10
Master_Log_File: mysql-bin.000049
Read_Master_Log_Pos: 107
Relay_Log_File: mysqld-relay-bin.000004
Relay_Log_Pos: 253
Relay_Master_Log_File: mysql-bin.000049
Slave_IO_Running: No
Slave_SQL_Running: No
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 107
Relay_Log_Space: 839
Until_Condition: None
Until_Log_File:
Until_Log_Pos: 0
Master_SSL_Allowed: No
Master_SSL_CA_File:
Master_SSL_CA_Path:
Master_SSL_Cert:
Master_SSL_Cipher:
Master_SSL_Key:
Seconds_Behind_Master: NULL
Master_SSL_Verify_Server_Cert: No
Last_IO_Errno: 2003
Last_IO_Error: error reconnecting to master 'replica@db01.mynetwork.net:3306' - retry-time: 10 retries: 86400
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 1
1 row in set (0.00 sec)
And finally this is the error I get running
# masterha_manager --remove_dead_master_conf --conf=/etc/mha/app1.cnf
Tue Sep 25 15:28:10 2012 - [warning] SQL Thread is stopped(no error) on
db02.mynetwork.net(11.22.33.2:3306)
Tue Sep 25 15:28:10 2012 -
[error][/usr/share/perl5/vendor_perl/MHA/ServerManager.pm, ln732] Multi-master
configuration is detected, but two or more masters are either writable
(read-only is not set) or dead! Check configurations for details. Master
configurations are as below:
Master db01.mynetwork.net(11.22.33.1:3306), dead
Master db02.db.cert.fronter.net(11.22.33.2:3306), replicating from
db01.mynetwork.net(11.22.33.1:3306)
Tue Sep 25 15:28:10 2012 -
[error][/usr/share/perl5/vendor_perl/MHA/MasterMonitor.pm, ln383] Error happend
on checking configurations. at
/usr/share/perl5/vendor_perl/MHA/MasterMonitor.pm line 298
Tue Sep 25 15:28:10 2012 -
[error][/usr/share/perl5/vendor_perl/MHA/MasterMonitor.pm, ln478] Error
happened on monitoring servers.
Tue Sep 25 15:28:10 2012 - [info] Got exit code 1 (Not master dead).
Is it a known issue? Any idea why this fails?
Original issue reported on code.google.com by m.je...@gmail.com on 25 Sep 2012 at 1:31
Original issue reported on code.google.com by
m.je...@gmail.comon 25 Sep 2012 at 1:31