Render messenger reporting?

[robcat2075]

Here's a current job running. Is this common?

 

Instance 0 got started first and is doing something but Render Server doesn't seem to know. But instance 1 appears to be reporting in correctly.

[robcat2075]

Here's another situation I'm running into.

 

The Render Slave has finished frame 29 and says it is "Waiting for next frame" It doesn't appear to have crashed. There are no errors reported anywhere yet.

 

Meanwhile Render Server was frame 30 ready to go but can't seem to get the Render Slave to take it.

 

Is there a cause/solution for this besides manually closing the render Slave and restarting it?

 

[Fuchur]

I dont think so. You will have to close the renderslave and reopen it.

I saw those bugs in the first releases of v16 alpha but they are no longer common. So I think it could be some trouble with the network-speed.

When my homeserver was running slow due to a ethernet-driver issue it happend all the time. Then I updated it and I never had a hanging renderslave again.

Maybe you can write a request to only wait a specified time for a new frame and if it doesnt receive one, it should reopen the renderslave automatically.

 

For now: Close the renderslave and reopen it.

You wont loose any data but the frame that was currently rendered by the slave. (I pressume you are rendering to image-sequences)

 

See you

*Fuchur*

[robcat2075]

I dont think so. You will have to close the renderslave and reopen it.

I saw those bugs in the first releases of v16 alpha but they are no longer common. So I think it could be some trouble with the network-speed.

 

It must be something other than network speed since the render server and render slave are on the same machine, right?

[HomeSlice]

I have run into that occasionally too. I don't know what causes it, but restarting the affected render messenger, like Fuchur said, seems to fix it.