Large number of self suspends (dispatches): <number> per command

There are a high number of task displacements in the database. The processing of long-running applications is terminated after a set time (XPARAM parameter _MAXRGN_REQUEST), to prevent other transactions from being blocked in the database (similar to timeslices in operating systems). To determine whether this is actually a problem, check the application profile. For example, complex searches in the data cache almost always dramatically increase the number of self-suspends. A high number of self-suspends is definitely indicative of a high number of long-running statements.

A task can execute a self-suspend when another task with higher priority changes from status waiting to status ready (only for XPARAM _MP_DISP_PRIO=YES).

For background applications, the number of self-suspends can be reduced by increasing the XPARAM parameter _MAXRGN_REQUEST. This can improve throughput because the system will now sequentially process the statements, but creates a disadvantage for short-running statements due to higher dialog response times.

If your analysis of the database application does not reveal any indication of complex statements, check whether the SQL statements are requiring much more data to be read than is actually required for processing (for example, due to tablescans or poor search strategies).