Example thread pool backlog entries from an uptime.log file:
2010-05-18 03:15:07,473 WARN (Launcher:117) - Thread pool backlog: Events in queue: 110 Available: 0 Busy:75 2010-05-18 03:15:07,474 WARN (Launcher:117) - Thread pool backlog: Events in queue: 111 Available: 0 Busy:75 2010-05-18 03:15:07,474 WARN (Launcher:117) - Thread pool backlog: Events in queue: 112 Available: 0 Busy:75
These warning messages indicate that Uptime Infrastructure Monitor's task scheduler is taking on new tasks faster than current tasks can be executed. These messages are generally not an indication of a significant problem if they occur intermittently and in short bursts. However, if these messages are sustained for extended periods of time, please contact uptime Support for guidance on potential system tuning.
There are various reasons why you may be seeing these messages:
- The serviceThreads variable in the uptime.conf file is set to a higher value than the connectionPoolMaximum parameter. Note that the connectionPoolMaximum value should be at least 10% higher than the serviceThreads value.
- Your hardware platform is saturated and can't execute tasks quickly enough.
- The average time required to complete a task is simply too long. If you have many monitors that consistently fail or reach their timeout setting, this can substantially impact the Uptime Infrastructure Monitor task list. uptime Support can provide further information and guidance on how to improve task execution performance and optimize the size of your task list.