Announcement

Collapse
No announcement yet.

Who was the fastest last week??

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    As I emailed to Martin earlier, I had problems with the seti.matroxusers.com site again... Trying to get a Delta WU over the past 6 days, got me the dreaded Internal Server Error.

    I thought it might be due to the fact that the software couldn't do this since 6 days ago, so I went back to 4 days, list 1-100, showing rank... Internal Server Error.

    3 days? Internal Server Error...
    2 days as in the original? Yeps, the error again... grrrrrrrrrr....
    1 day then, please? That worked !!! Great !!!

    I tested this in Netscape 4.73 at first, then started up IE5 on it's own, and tested again.

    Same errors for all the pages if you went bigger than 5 the first time. In the end only 1 page, list 1-100, rank would show...

    Weird?

    Maybe something to look into, Ant & Martin.

    Jord.

    ------------------
    "What goes up, must come down"
    Jordâ„¢

    Comment


    • #17
      Right,

      I think I may have found the cause for the internal server error. If any of you still get one (on the murc-page, not the other two), please let me know (in here, or at ees@casema.net).

      Martin

      Comment


      • #18
        How about three sortings.

        1, Fastest (the current one)
        2, Volume (the new request)
        3, Hours (after all, this is the one that proves dedication)

        chuck
        Chuck
        秋音的爸爸

        Comment


        • #19
          Regarding the "low on resources" thingy...

          Seti (at least the CLI version) has a handle leak when running under Win98. It leaks about 1-2 handles pr. second on my machine so if it isn't stopped at least once pr. WU it will eventually slow your machine to a crawl...

          Just thought I'd let you know.

          Comment


          • #20
            I do suspect that one. I have not done much with the computer but recourses down to 19% :-(

            Any idea how to investigate (and stop) this leakage??

            Annyone else with this problem??




            ------------------
            Handle: WinWiz
            Asus P3B-f PIII 558 G400Max 256Mb 57 Gb SbLive Plextor 12432 & Ultraplex 40max Hp 970 Cxi Hp Scanjet 3c Iiyama Vm450
            Handle: WinWiz
            Asus A7V133 Thubderbird 1400 Geforce3 512Mb 106 Gb SbLive Plextor 12432 & Ultraplex 40max Hp 970 Cxi Hp Scanjet 3c Iiyama Vm450 Pinnacle Dv500

            Comment


            • #21
              Stopping and starting seti between each WU is ok for me... it still leaks a lot of handles, but not quite enough to slow my pc.

              I haven't got a clue how to fix the leak, and I suspect it might be in Win98's NT virtual machine (yes, seti is run by an NT kernel in a virtual machine in Win98). There is no leaking on NT.

              Comment


              • #22
                Kruse

                >>>>Seti (at least the CLI version) has a handle leak when running under Win98. It leaks about 1-2 handles pr. second <<<<

                errm ,I've no idea what your talking about!

                Could you enlighten me please?


                ------------------
                I guess I won't recruit for my team here TA

                Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, Asteroids@H, LHC@H, Skynet POGS.

                Main rig - Q9550 @3.6 GHz, HD 5850 (Cat 13.1), 4GB DDR2, Win 7 64bit, BOINC 7.2.42
                2nd rig - E5200 @3.73 GHz, GTX 260 c216, 4GB DDR2, Win XP, BOINC 7.2.42

                Comment


                • #23
                  Hmm... TA needs help, eh? I guess this might get me kicked from MURC, but here goes anyway...

                  Get TaskInfo2000 if you don't already have it from http://www.iarsn.com/download.html#TaskInfo

                  Then start TaskInfo and select your Seti task (seti.exe or whatever you TA guys like to call it ;-) Then look at the handle count in the lower right corner... tick, tick, tick...

                  In case you don't know: Handles are pointers to various Windows resources. A handle leak means some resources are not deallocated correctly... But as I mentioned the bug is probably in the NT VM and not in seti.

                  ------------------

                  Comment


                  • #24
                    Thanks Kruse ,its just for me 'cos I'm clueless about that kinda thing!

                    >>>>In case you don't know: Handles are pointers to various Windows resources<<<<

                    That's a start ,could you expand on that?

                    Thanks

                    (BTW I did already have taskinfo 2000 )
                    Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, Asteroids@H, LHC@H, Skynet POGS.

                    Main rig - Q9550 @3.6 GHz, HD 5850 (Cat 13.1), 4GB DDR2, Win 7 64bit, BOINC 7.2.42
                    2nd rig - E5200 @3.73 GHz, GTX 260 c216, 4GB DDR2, Win XP, BOINC 7.2.42

                    Comment


                    • #25
                      Hmm... this could turn into "Windows Programming 101"...

                      A pointer is an address, right! (that's about as basic as I'm willing to get ;-)

                      Every time a program allocates an object it gets a pointer to that object. If the object is some kind of resource (window, document, brush, cursor, bitmap etc.) the pointer is called a handle.

                      If the program runs a loop like this:

                      Begin loop
                      Allocate a resource
                      Do something with it
                      Deallocate the resource *
                      End Loop

                      but forgets the step marked with a *, it has a handle leak. That means Windows thinks the program uses more and more resources even though it only uses one (or a few) at a time.

                      That's a major problem, especially in languages like C++ (which Windows itself is written in) because the programmer has to do all the accounting and remeber to deallocate all objects when they are no longer used.

                      Hope that enlightened you a little more Assim. But it's a vast subject, and I'm not really an expert...

                      If I'm wrong please correct me. And if anyone has a good url with more information, this thread would be a nice place to post it in ;-)

                      Comment

                      Working...
                      X