Bug #8282 Replication stops on change made in .cnf file and restart

• Submitted: 3 Feb 2005 13:13
• Modified: 4 Feb 2005 11:07
• Reporter: Sumit Kumar Roy
• Status: Not a Bug
• Category: MySQL Server: Replication
• Severity: S1 (Critical)
• Version: 4.0.17
• OS: Linux (Linux 7.1 Kernal 2.4.26)
• CPU Architecture: Any

[3 Feb 2005 13:13] Sumit Kumar Roy

Description:
We have 2 instances of MySQL running on a machine, one instance as Master and the other in Slave mode. 

In the beginning we did not mention the relay log file name in the Slave configuration file. The replication was happening properly. 

Now, we changed the slave configuration file and added the following lines:
relay-log=/usr/local/archivedb/data_s/standby_rlog
relay-log-index=/usr/local/archivedb/data_s/standby_rindex.index
log-bin=/usr/local/archivedb/log_s/db2_log_bin
log-bin-index=/usr/local/archivedb/log_s/db2_log_bin.index

After the above change both Master and Slave were restarted. Now the Slave stopped replication.

How to repeat:
Follow the steps in bugs description

[4 Feb 2005 7:32] Victoria Reznichenko

Did you have any error in the error log file on the slave?
Check that you removed relay-log.info file.

[4 Feb 2005 8:16] Sumit Kumar Roy

No removal of files were done in the slave directory.

[4 Feb 2005 11:07] Victoria Reznichenko

relay-log.info file is updated by SQL thread and keeps info about relay log files, position etc. If you don't remove this file, it still has old info.