VRay Next GPU Distributed Buckets Drop Out...

[joelly3d]

Using both CPU and GPU buckets I've noticed that when rendering a sequence, different machines tend to stop rendering for no reason. Is there something I can do to make it more stable? I have 10 40core machines with 1660's in them. Works amazing on the first few frames, then they seem to drop out randomly. It's a heavy scene, 15M polys, but not alot of texture data. I am on 2.1, not the nightly.
Update: Seems more stable with GPU only. Perhaps there is a some choking that is happening with CPU+GPU combo that causes them to both stop? I hate to waste all that horsepower and not utilize it.
Any thoughts?
-Joel

[joelly3d]

Update: My host machine had a P5000 and an RTX 5000 in it. I switched to and RTX4000/RTX5000 and it seems more stable rendered all night without any failures. Not sure if this is possible, but wanted to mention it. Maybe things are just always more stable if hardware families are matched?

[Martin.Minev]

Hello,

Both RTX and GTX cards should work together in a distributed render. Could you send us the scene for investigation to support@chaosgroup.com. If you could Include the VRayLog.txt file from the temp folder (%temp%) it will help us troubleshoot. Please also mention this thread.

[joelly3d]

Opening a fresh copy of max, it didn't do it again. So perhaps there is a choke or backup somehow in Max itself? (no shock there). I will save logs next time I run into the occurrence and kickem over. Thanks Martin!