Increasing Thread Pool Size

It may help to increase the Thread Pool size if:

  • The average server-side processing duration for activities increases.
  • The number of activities entered/activated per day increases.
  • The number of new process instances per day increases (because this will typically increase the number of activities entered/activated per day).
  • The average number of activities per running process instance increases (i.e. if you start deploying much more complicated process models than you were, it increases the likelihood of concurrent processing).
  • The average duration (from Start to Complete) of your process instances increases (because this may increase the number of activities entered/activated per day). This is especially true for processes whose duration is determined by the number of times the process instance "loops" back and re-enters an activity earlier in the process model. It may not be relevant for processes whose duration is primarily determined by the durations of the Manual activities in the process.
  • Any of your commonly-used processes contain activities that require minutes (or longer) of server-side processing (as described above), as this significantly increases the likelihood of concurrent processing.
  • Your environment is prone to intermittent periods of much-higher-than-average load during certain times of day (or week, month, etc.). For example, if your process instances are often started nearly simultaneously (e.g. at the start of the work day or shift, or because of certain events like the upload of a large batch of files), or you have many activities that "time out" simultaneously (e.g. at the end of the month), then it may make sense to increase the Thread Pool size to take the higher-load periods into account.
  • Your server, network, or database is prone to intermittent "slowdowns" (i.e. periods of slower-than-normal processing). E.g. This could be due to scheduled services running in the background at specified times of the day, or could be caused by periods of high server load caused by non-AgilePoint sites or applications (on a non-dedicated server).
  • Your server has plenty of memory (i.e. RAM), but not a lot of processing speed (i.e. CPU speed). In other words, activities will take longer to process on a slow machine.