The RPE Web Services provides a queue mechanism for remote document generation. The mechanism was introduced to avoid the RPE Web Services from causing Out of Memory errors since all remote jobs were executed in the same 32 bit server process which was limited to less than 2Gb of memory.
Update 2013.12.10: there is a typo in the current Infocenter as the flag is com.ibm.rational.rpe.docgen.concurrentlimit (double r) and not com.ibm.rational.rpe.docgen.concurentlimit ( single r)
But with RPE 1.2 Web Services being redesigned from ground up each document generation job is run as a separate 32 or 64 bit process which means that memory is no longer an issue. However the queue is still necessary as CPU is now the limiting factor.
To control the queue size in RPE 126.96.36.199 and older you had to modify the “maxConcurrentProcesses” entry in the rpeconfig.xml file. That entry no longer exists in the config file of PE 1.2 and newer. To control the queue size in RPE 1.2 and newer you need to define the “com.ibm.rational.rpe.docgen.concurrentlimit” (double r) property in the App container that runs the WebService.
If the property is not defined than the RPE Queue Size defaults to 7. But the actual value to use depends on server hardware and data source type (ex: running DOORS reports is more expensive CPU wise as a new DOORS instance has to be created) so we recommend to set it based on your actual configuration and needs. A value of 4-5 is a good starting point for most configurations.
More details on all the RPE Remote flags: