This option enables you to adjust for any delay in the service call from the Process Engine or WmTaskClient to the Task Engine caused by the effects of network latency. You can use this setting to specify any additional delay in the service call when the Task Engine is under heavy load and the actual time it takes to create a task is longer than the socket timeout configured in WmTaskClient. In such conditions, even though a socket timeout occurs, the Task Engine might still proceed with the task creation instead of deleting the task.

For example, suppose that the Process Engine invokes WmTaskClient to queue a task that is part of a process and the socket timeout is set at the default value of 60000 milliseconds. If the additional delay in network latency is 50 milliseconds, the total timeout (that is, from the time the Process Engine issued the task creation request to the response from the Task Engine) would be 60050 milliseconds. To account for this, you can set -DqueueTask_ws.delay.time to 50 milliseconds, which would adjust the Task Engine timeout to 59950 milliseconds, thus returning the total timeout to 60000 milliseconds.

The default value of -DqueueTask_ws.delay.time is 20 milliseconds. When setting the parameter, Software AG recommends specifying a value that is greater than the default value.