Details

    • Type: Bug Bug
    • Status: Closed
    • Priority: Major Major
    • Resolution: Fixed
    • Affects Version/s: 0.1
    • Fix Version/s: 2.0-alpha1
    • Component/s: Core
    • Labels:
      None
    • Environment:

      windows
      NUMBER_OF_PROCESSORS=2
      OS=Windows_NT

      Description

      got code from trunk ran mvn.bat clean install

      ... see attached

      i cant attach files here need special rights for that?

      see zip at http://sel2in.com/pages/log4j/2012march21surefire-reports.zip

      march21IST.txt has console out

        Activity

        Hide
        Ralph Goers added a comment -

        Closing - will make sure that the recommendation is to always use -server.

        Show
        Ralph Goers added a comment - Closing - will make sure that the recommendation is to always use -server.
        Hide
        Ralph Goers added a comment -

        FWIW - I discovered that adding -Dmaven.surefire.debug="-server" to the command line for SimplePerfTest yields performance similar to OS/X and Linux. The implication is that Log4j2 is benefitting greatly from the Hotspot compiler (27 nanoseconds per iteration with the default and 2 nanoseconds per iteration with -server). This probably also explains why, when I tried to to profiling with YourKit making changes to eliminate what YourKit was finding as hot spots made absolutely no difference as HotSpot was already eliminating them. However, making the changes manually might bring the performance in -client mode closer to -server.

        Show
        Ralph Goers added a comment - FWIW - I discovered that adding -Dmaven.surefire.debug="-server" to the command line for SimplePerfTest yields performance similar to OS/X and Linux. The implication is that Log4j2 is benefitting greatly from the Hotspot compiler (27 nanoseconds per iteration with the default and 2 nanoseconds per iteration with -server). This probably also explains why, when I tried to to profiling with YourKit making changes to eliminate what YourKit was finding as hot spots made absolutely no difference as HotSpot was already eliminating them. However, making the changes manually might bring the performance in -client mode closer to -server.
        Hide
        Ralph Goers added a comment -

        I was able to recreate this in a Windows VM and was able to address all the unit test failures. However, I noticed that the performance on Windows in the VM is considerably slower than in OS/X or even a very old Linux machine. I need to dig up an old Windows machine and see if it performs better running natively.

        However, as the issues raised here should be fixed it would be better to deal with that in a separate issue.

        Show
        Ralph Goers added a comment - I was able to recreate this in a Windows VM and was able to address all the unit test failures. However, I noticed that the performance on Windows in the VM is considerably slower than in OS/X or even a very old Linux machine. I need to dig up an old Windows machine and see if it performs better running natively. However, as the issues raised here should be fixed it would be better to deal with that in a separate issue.

          People

          • Assignee:
            Ralph Goers
            Reporter:
            Tushar
          • Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Development