Recently, a customer asked us:
Why would heavy disk IO cause the Tungsten Manager and not MySQL to be starved of resources?
For example, we saw the following in the Manager log file tmsvc.log: