Dominik Rau wrote:
I've got some really annoying sync problems (no, not the changelist
this time ; )) here with my cluster (animation hangs shortly sometimes
in unpredictable intervals) and I'm quite sure that they are related
to a badly configured network, especially wrong routing (I think).
Please, could anybody give me some hints about multicast under linux
(debian sid with 2.6 kernel in my case, but that shouldn't make that
much difference)?
Maybe I should explain why I think that it's routing related:
My client starts all servers via ssh remotely, also the one on localhost
(I'm rendering on the client, too). If I run the client and one server
on the localhost only and put my default route on the local loopback
device, it works like a charm. If I route the default to eth0 like
above, all smoothness is gone. Even more funny: If I have less objects
in my scene, it's getting worse (faster frame rate but more bad syncing
I think). It also makes no difference if I run the client on a dedicated
cluster node without rendering on it. Can't see why my dedicated and
switched 1000MBit network should be a bottle-neck here (nload shows me
about 8MByte traffic per second).
So, it looks like bad routing to me. Any other opinions?
Regards,
Dominik
-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users
