Bakburner workflow suggestions for 2 users

[Streetwise]

I think I've asked this before perhaps, but as a single user, I have a particular setup with 4 nodes, of which BB server is running on one of them. I pass jobs off to it and life is good.

I have a client who has two users on workstations, and 4 nodes. The IT dept has server installed on all 4 nodes and manager running on another internal machine somewhere. The two workers are always frustrated because "the farm is down" frequently. I don't have a complete list of the specific issues, but it would seem like running server on all 4 nodes isn't really necessary is it? It would seem like you could have two instances running and then each user would always send a job to the same server. The other two non-server nodes would just help out with whosever job? How would you set it up?

[Morne]

Is this for animation?

Does each user want 2 nodes specifically for himself or sometimes all 4?

In backburner you can set groups. You can make group A (with server 1 and 2) and group B (with server 3 and 4)

When user 1 submits the BB job, he uses Group A. When user 2 submits the BB job, he uses group B. That will stop them fighting over each other's server's, however it also means a waste for the times when one of the users aren't rendering in which case 2 servers will sit idle.

Or you can simply manage the job via BB monitor. In monitor you can assign or remove servers on the fly during a job. So if user1 is rendering on server 1 and 2 and 3 and 4 is not used, he can just add it. If the other user then wants to render his job, he can just remove 3 and 4 from current job and assign it to his job

[Streetwise]

No animation at all. Each designer has their own projects that they may be working on. In general each always has something going on render-wise. So with Groups, does Server need to be running on all 4 nodes? I would only think just one node per group, correct? Also, in the V-Ray DR settings, do you need to have ALL servers listed?

In the case of adding servers on the fly: So you would have Server running on 1 and 2. User 1 submits to Server 1, User 2 submits to Server 2. User 2 didn't show up for work that day, so User 1 can then add 2,3 and 4 to their job? In the V-Ray DR settings, would each user only have their assigned server listed? Does that make sense?

Thanks

[kosso_olli]

We use Backburner here, too. We all send our jobs to the same server on our renderfarm. This server then starts to render, and then DR kicks in with all the other render nodes.
All you have to do is enable DR before you send the job to Backburner.
This is first come, first serve though. The queue might get quite long, but we never had any problems with this. Backburner is better then most people think.

[Streetwise]

We use Backburner here, too. We all send our jobs to the same server on our renderfarm. This server then starts to render, and then DR kicks in with all the other render nodes.
All you have to do is enable DR before you send the job to Backburner.
This is first come, first serve though. The queue might get quite long, but we never had any problems with this. Backburner is better then most people think.

This is EXACTLY how I roll. But I'm a single user, so it's easy. I suspect though even with a two person team, it wouldn't be hard to negotiate and prioritize jobs. I'm also recommending that they include their own workstations in DR so that at night, they can turn on Spawner and take advantage of two more nodes. I know they aren't doing that now.

[thablanch]

Was Backburner installed using a <clone machine> software ?
Had similar issues where, it was not using all the machines as the machines were installed from 'an image' and it caused BackBurner problems, really similar to what you are talking about.
I did reinstall Backburner, and made sure that the server names were different, and the problem was solved. All the nodes were participating

[Streetwise]

Was Backburner installed using a <clone machine> software ?

Probably. Large company with IT dept. That's generally how they roll. Anyhow, I was able to go in and set it up the way I'm familiar with and it all seems to work now. Basically, Manager is off on some server they have, and Server is set up on one of the 4 nodes for them to send jobs too. Using Monitor, they can then re-prioritize jobs in the queue. So far all nodes are participating, plus their own workstations when they go home at night and turn Spawner on. The next step would be to set up "groups" if this workflow doesn't seem to work for them.