Bug #36450 Replication failing after upgrading to 5.0.51b

• Submitted: 1 May 2008 10:22
• Modified: 13 Aug 2008 14:45
• Reporter: Ashley Unitt
• Status: Closed
• Category: MySQL Server: Replication
• Severity: S1 (Critical)
• Version: 5.0.51b
• OS: Windows (Server 2003)
• CPU Architecture: Any

[1 May 2008 10:22] Ashley Unitt

Description:
We have a couple of servers set up with master<->master replication that had been working without issues on 5.0.45.

This morning we've upgraded them to version 5.0.51b and replication seems to fail on start-up.

SHOW PROCESSLIST gives:

170 replication patty.development.newvoicemedia.com:3282  Binlog Dump 1073 *** DEAD ***                                                                                                                     
171 root        MAJORCLANGER:5285                         Query       0                  show processlist                                                                                   
643 system user                                           Connect     12   *** DEAD ***                                                                                                                     
644 system user                                           Connect     11   *** DEAD ***                                  

The error log gives:

080501 10:58:01  InnoDB: Started; log sequence number 0 1361851356
080501 10:58:02 [Note] C:\Program Files\MySQL\MySQL Server 5.0\bin\mysqld-nt: ready for connections.
Version: '5.0.51b-community-nt-log'  socket: ''  port: 3306  MySQL Community Edition (GPL)
080501 10:58:02 [Note] Slave SQL thread initialized, starting replication in log 'selma-bin.000035' at position 98, relay log '.\patty-relay-bin.001141' position: 235
080501 10:58:02 [Note] Slave I/O thread: connected to master 'replication@selma:3306',  replication started in log 'selma-bin.000035' at position 98
080501 11:04:52 [ERROR] Error reading packet from server: Lost connection to MySQL server during query ( server_errno=2013)
080501 11:04:52 [Note] Slave I/O thread killed while reading event
080501 11:04:52 [Note] Slave I/O thread exiting, read up to log 'selma-bin.000035', position 98
080501 11:04:52 [Note] Error reading relay log event: slave SQL thread was killed

There should have been no configuration changes from when we upgraded from 5.0.45 where the replication was working.                                                                                         

How to repeat:
Restart the server or issues a SLAVE START;

[7 Aug 2008 9:16] Susanne Ebrecht

Many thanks for writing a bug report. Which exact query occurs the crash?
How did you upgrade the slave?

[13 Aug 2008 14:45] Ashley Unitt

We upgraded the slave just by running the standard Windows install package.

Unfortunately we've had to recommission the server that we were running on so I'm no longer able to replicate the issue so I've decided to close it.

Apologies

Ashley