Bug #43370 | Replicated epochs fail to end up in slave cluster binlog | ||
---|---|---|---|
Submitted: | 4 Mar 2009 13:39 | Modified: | 24 Mar 2009 17:21 |
Reporter: | Lars Torstensson | Email Updates: | |
Status: | Closed | Impact on me: | |
Category: | MySQL Cluster: Replication | Severity: | S3 (Non-critical) |
Version: | mysqlcom-5.1.31-ndb-6.3.22-telco | OS: | Linux |
Assigned to: | Tomas Ulin | CPU Architecture: | Any |
[4 Mar 2009 13:39]
Lars Torstensson
[11 Mar 2009 13:00]
Tomas Ulin
note that log-slave-updates=1 must be set on master-21. please verify that this is the case. BR, Tomas
[11 Mar 2009 13:19]
Lars Torstensson
log-slave-updates=1 is verified on master-21.
[11 Mar 2009 13:44]
Tomas Ulin
can't reproduce: I do cd mysql-test edit suite/rpl_ndb/my.cnf -> add log-slave-updates to [mysqld] section ./mtr --start rpl_ndb_basic will create 2 cluster with 2 mysqld on one side and 1 on the other slave> change master to master_host='localhost', master_port=12503, master_log_file='slave-bin.000001', master_user='root'; slave> start slave; master> create table t1 (a int key) engine ndb; master> insert into t1 values (1); slave> show binlog events; slave1> show binlog events; both show the same contents... alias slave='../client/mysql test -u root --socket=var/tmp/mysqld.1.1.sock --prompt='\''slave> '\''' tomas@poseidon:~/mysql_src/mysql-5.1-telco-6.3/mysql-test$ alias master1 alias slave1='../client/mysql test -u root --socket=var/tmp/mysqld.2.1.sock --prompt='\''slave1> '\''' tomas@poseidon:~/mysql_src/mysql-5.1-telco-6.3/mysql-test$ alias slave alias master='../client/mysql test -u root --socket=var/tmp/mysqld.1.slave.sock --prompt='\''master> '\'''
[13 Mar 2009 13:28]
Lars Torstensson
Even I can't reproduce now... Think I have realized my mistake. The mysqld servers were not properly reinstalled, but that is covered in an other issue with a customer. Please close this /lars
[17 Mar 2009 18:35]
Andrew Hutchings
Re-opened open customer request