Help with errors on MySQL replication

The following is happening on two non-production Fedora Core 11 machines...


Unable to find any info regarding patching the Archive engine into a standard yum install, I removed the std package and compiled 5.1.42 from scratch.

All seemed fine, but slaves replicating off the back of this machine crashed. After a bit of investigation I got to the following repeatable error.

Start MySQL, run any query that writes to the replication log, then stop MySQL.

Run mysqlbinlog on the newly created replication log file and get something like;


#100126 17:18:36 server id 1 end_log_pos 106 Start: binlog v 4, server v 5.1.42-log created 100126 17:18:36 at startup
ROLLBACK/*!*/;
BINLOG '
7CNfSw8BAAAAZgAAAGoAAAAAAAQANS4xLjQyLWxvZwAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAADsI19LEzgNAAgAEgAEBAQEEgAAUwAEGggAAAAICAgC
'/*!*/;
ERROR: Error in Log_event::read_log_event(): 'Found invalid event in binary log', data_len: 94, event_type: 2
ERROR: Could not read entry at offset 106: Error in log format or read error.



If I don't do an update query (i.e. just start and stop MySQL with no transactions) the binlog is marked as OK by mysqlbinlog.