Revision 47 crash, Revision 51 upgrade, now getting error 1236

MySQL Revision 47 went away this morning with nothing in the error log except this:

100702 09:31:19 mysqld_safe Number of processes running now: 0
100702 09:31:19 mysqld_safe mysqld restarted

We brought mysql up again, checked all tables, then brought it down and upgraded to Revision 51. Now our slaves will not start, reporting:

Last_IO_Error: Got fatal error 1236 from master when reading data from binary log: ‘Could not find first log file name in binary log index file’

The binary logfiles haven’t moved, permissions haven’t changed, and if I FLUSH LOGS on the master, it properly creates a new binary file in the expected place and updates the index file accordingly. The first line in mysql-bin.index is:


and that file exists and is parseable by mysqlbinlog without errors. We’re at a complete loss.

Hi eamondaly,

This might probably be late, but how does your ‘SHOW SLAVE STATUS \G;’ look like?

To follow up on this, the error message was a little misleading. When MySQL crashed, it didn’t log the current binary file to the index, so there was a file missing in the middle of the list. I was able to add the name to the index file and FLUSH LOGS, and the slaves recovered.