1 2 Previous Next 24 Replies Latest reply: Jul 16, 2010 12:41 AM by 807557 Go to original post RSS
      • 15. Re: causes of jitter
        807557
        Note you can use the com.sun.rtsjx.SteadyMode class's enterSteadyMode() and exitSteadyMode() to put markers in the log produced by LogJitterProneEvents, so that you can see which jitter related events happen within specific sections of the program (of course this is time based so all events in all threads will be in the delineated section of the log)

        David Holmes
        • 16. Re: causes of jitter
          807557
          Hi,

          I used SteadyMode class's enterSteadyMode() and exitSteadyMode() and there are around 6500 patching calls happening between these two markers. Jitter in case of first order is not this high if I use Hotspot 1.6 so I can say that this problem is caused by patching calls only. Also now we are doing all pre-initializing/compiling through mechanism(vm switches at command line) provided by RTS.

          Can we know for sure why these many patching calls are happening during first order.

          Prashant
          • 17. Re: causes of jitter
            807557
            Are you using reflection in your application?

            Also just a sanity check but you do execute in a real-time thread when using java RTS, don't you?

            David Holmes
            • 18. Re: causes of jitter
              807557
              dontwantto wrote:
              Are you using reflection in your application?
              Yes we are using refection in application though not very extensively. Patching calls are occurring for methods not having any reflection code.
              >
              Also just a sanity check but you do execute in a real-time thread when using java RTS, don't you?
              yes we are executing javax.realtime.RealtimeThread

              Prashant
              • 19. Re: causes of jitter
                807557
                I tried running the application without using reflection, patching calls are still happening.
                • 20. Re: causes of jitter
                  807557
                  I've done some experimentation and I also see patching occurring during steady-state mode with ITC being used. I don't know why this is the case and I've opened a bug report for it.

                  Sorry.

                  David Holmes
                  • 21. Re: causes of jitter
                    807557
                    Thanks David,

                    Is there any plan to release fixed RTS in near future?

                    Also can we extend our 90 day evaluation as this bug has taken a big chunk of our evaluation time. And we still have quite a lot performance testing and code updating to do before we can determine if RTS will be useful to us.


                    Prashant
                    • 22. Re: causes of jitter
                      807557
                      Prashant-Goel wrote:
                      Also can we extend our 90 day evaluation as this bug has taken a big chunk of our evaluation time. And we still have quite a lot performance testing and code updating to do before we can determine if RTS will be useful to us.
                      Please just re-download Java RTS again to continue your evaluation.

                      David Holmes
                      • 23. Re: causes of jitter
                        807557
                        Hi David,

                        Is there any ETA reagarding this bug fux?

                        Thanks,
                        Prashant
                        • 24. Re: causes of jitter
                          807557
                          Sorry I have no additional information.

                          David Holmes
                          Oracle
                          1 2 Previous Next