Announcement

Collapse
No announcement yet.

More odd spawner behaviour

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

  • More odd spawner behaviour

    Running a 4k render using DR to 6 render nodes, 3 of which have 16GB RAM and 3 have 32GB. The scene only seems to be using around 5GB when rendering. The lightcache and irradience map seemed to calculate just fine, but whilst rendering the buckets, the nodes seem to randomly drop out (warning: Render host slave5 (192.168.0.5) is not responding) and then moments later, pick up the scene again. This seems to be happening to all of the machines.

    What gives? Is there an error log I can send to you to investigate?
    Kind Regards,
    Richard Birket
    ----------------------------------->
    http://www.blinkimage.com

    ----------------------------------->

  • #2
    Yes it's very weird indeed. At one of my sites we also have 6 slaves. For ALL my render jobs to these slaves, the one machine stops and starts the whole time and never actually render. (The slaves are identical clones and just network name and ip changed). However another user always renders just fine with that slave that drops for me...
    Kind Regards,
    Morne

    Comment


    • #3
      Have you tried closing the dr spawner and then starting again? I feel like I've had to do that a couple times to clear cache & freshen up the node that was acting up. I think something goes wrong at times from quickly quitting renders and starting them again with the DR spawners -
      Brendan Coyle | www.brendancoyle.com

      Comment


      • #4
        Originally posted by cheerioboy View Post
        Have you tried closing the dr spawner and then starting again? I feel like I've had to do that a couple times to clear cache & freshen up the node that was acting up. I think something goes wrong at times from quickly quitting renders and starting them again with the DR spawners -
        Yes. Done this and even restarted the machines. Its bloody annoying.

        I am now investigating going back to strip rendering through backburner with pre-calc'd solutions. Haven't done that since about 2008!

        So far I have discovered that we cannot save to EXR using this method. If we do, it only saves one strip. Is this a vray bug or a max bug?

        Also, it looks like we must disable the vray frame buffer if we use render elements (which we do all the time) as the element strips fail when trying to put them back together. Is this a vray bug or a max bug?

        It also seems that regardless of the file save name I put in the render dialogue box, the render elements do not update their names. Is this a vray bug or a max bug?

        Apart from these errors, I can at least get render out...at the moment. We seem to have taken a number of steps backwards in recent months.
        Kind Regards,
        Richard Birket
        ----------------------------------->
        http://www.blinkimage.com

        ----------------------------------->

        Comment


        • #5
          We have to test all those things in order to find out who's responsible for the bugs.
          Would it be possible to send us a sample scene with this setup - it would be easier to work directly on your file instead of wondering how you have created it.
          You could delete everything from the file , we will only need the settings.

          About spawners dropping - please make sure that "restart slave on render end" is checked - and also ensure that you have the same version installed on all render slaves and workstations.
          Svetlozar Draganov | Senior Manager 3D Support | contact us
          Chaos & Enscape & Cylindo are now one!

          Comment


          • #6
            Originally posted by svetlozar_draganov View Post
            We have to test all those things in order to find out who's responsible for the bugs.
            Would it be possible to send us a sample scene with this setup - it would be easier to work directly on your file instead of wondering how you have created it.
            You could delete everything from the file , we will only need the settings.

            About spawners dropping - please make sure that "restart slave on render end" is checked - and also ensure that you have the same version installed on all render slaves and workstations.
            What email address should I send it to?

            I always have 'restart on end' checked.
            Kind Regards,
            Richard Birket
            ----------------------------------->
            http://www.blinkimage.com

            ----------------------------------->

            Comment


            • #7
              Please send it to support@chaosgroup.com and include a link to this thread.
              Thank you very much in advance.
              Svetlozar Draganov | Senior Manager 3D Support | contact us
              Chaos & Enscape & Cylindo are now one!

              Comment

              Working...
              X