DR: machine released because there is no more pending work

[MANUEL_MOUSIOL]

So I was rendering with two machines with the latest official Vnext GPU build and everything seemed fine - both machines rendering with CPU and GPU in hybrid.
But then the distributed render slave machine stopped (although it still seemed to think it was rendering per the logs) and I got the message in the master machine that the other machine was "released because there is no more pending work".
One moment i thought it is funny that the other machine was so lazy it just decided it could go on vacation, faking that work was done.
But I am (of course) on a deadline and now I am stuck with one machine and a much longer render time, which will not finish by today.
Any idea, what is going on here and how to solve this?

[matanov]

Can you send me the logs of both the client and the server machines?

[vlado]

If you restart the V-Ray spawner on the other machine, it should be able to join the rendering again in any case.

It would be very helpful to get the logs though, if you keep them.

Also, is this with bucket or progressive rendering?

Best regards,
Vlado

[MANUEL_MOUSIOL]

This is with buckets.
i have to do another test. Don't think the logs are still there .

[grantwarwick]

Getting the exact same problem as described here.

[matanov]

Getting the exact same problem as described here.

I really can't help you without additional info like logs, scenes and repro steps.

[marksmanguy]

I am getting this too. Using the latest nightlies. Happens in every scene. Sending you my vray log.

[marksmanguy]

What I was getting wrong was that I didn't have the GPU render server launched on the slave. I was used to CPU DR with just the spawner.

Hope that helps

[joelly3d]

I have this problem as well.
Slave has VRayNext GPU Render Server running,
...then it works for a bit. About halfway through the render the buckets stop
...then host says machine says "released because there is no more pending work"
...then the slave says "waiting for render requests".

Odd. It's not consistent. Sometimes no issues, sometimes it won't let me use all my resources like this. Have 10 render nodes, for this particular 8k print render, 2 stopped with that error for no reason. Note that closing and restarting the GPU Render Server didn't kick it back into the mix as I would expect it to.

[joelly3d]

The longer the render went, the more drop out... just way earlier than they should. Perhaps the estimated remaining buckets calculation has an error?

[kubbek]

I have the same problem - trying to render with 4 slaves but two get dropped with the "released because there is no more pending work" information in the v-ray messages window. I attache log files

Edit: After a while all servers get dropped

[joelly3d]

kubbek ....just type %temp% in windows explorer and it'll drop you there.

[kubbek]

Just tested and the slaves are dropped as well on bucket as on progressive rendering

[meanadam]

Was this solved? Having the same problem today. Newly installed Max 2021, if it makes a difference.

Edit: I started the spawner manually and it works. NOT as a service.... Previously running the spawner as a service fixed my DR issues as there were user permissions or something due to the user profiles blocking the machines.

[joelly3d]

No, I still have issues like this. Seems random. I have a feeling it is because I am on a machine with more than 64 cores.