views:

105

answers:

1

I have a MySQL database that has on the order of 400 million innodb_data_pending_reads and innodb_pending_writes. My other databases are consistently at or near zero, so I noticed this large outlier.

What can cause this situation?

What adverse affects can it cause?

How can I troubleshoot the situation to bring this down?

A: 

After discussing this with MySQL enterprise support it would appear this is a known issue with the statistics reporting in version 5.0.18, which that particular server is running. The server does not appear to be suffering, so this is simply a false emergency.

The fix is to: 1) upgrade the server version 2) use this command to get more reliable InnoDB stastics: Iostat -x -a

Chris