Everything related to Maxwell network rendering systems.
i'am an italian Architect and a i am new in this forum. I want ask you if is possible with network monitor to render more scenes in same time.
Now when i start a cooperative jobs this start but again when i start the second cooperative jobs i see the written "pending".
Is possible to change this?

Thanks to all
Hello, and welcome to the forum.

The only way to render several jobs at once is not to use cooperative mode, just use "Single job" mode. This will add the scene to the first available computer. Then each computer renders a scene. But there is really no time saving in running several jobs at once on separate computers, compared to 1 job on all computer which will naturally finish faster, then move on to the next render, ie:

5 jobs on 5 individual computers at the same time (single mode), will take the same amount of time as 1st job x5 computers, followed by 2nd job x 5 computers etc. (coop mode).

If it's for quickly getting an idea of what several scenes look like, just specify a lower SL, say 9. Let the queue finish, you examine the renders, the ones you wish to let render for longer, simply add the jobs back into the queue, Monitor will detect there are already MXI files in the specified Save location with the same name and ask if you wish to resume the job, or start from scratch. So you choose to resume it, and they will continue from SL 9 to whatever the new SL you chose.

This is a nice way of quickly getting "previews" of several renders, because reaching SL 9 in coop mode, will be fast.
Sorry to tag on to this thread, I have a question related though. I must not be understanding the Maxwell Netrender Monitor very well. It looks like when a Cooperative Job is underway that , in my case when I have 3 machines running to a SL of 17, that all three machines are contributing at various percentages to get to the SL of 17. But when I add up the SL of the three machines in the progress column, it is far above the 17 with each machine looking like it worked to an SL of 14.2 to be added together to make the total SL of 17.

Is this because it is not a linear scale between SL points and that you need the machines to each go to that level to get to a single SL of 17? Or would it be more efficient if I have 3 renders to send each render to a machine by itself as a Single Job. Would that get me to all three renders being completed faster than each render being worked on in the Cooperative Job mode one after another?

Hope that made sense..
Well to keep things simple, all that network rendering does is facilitate the distribution of jobs. Whether you start manually 3 renders on your 3 computers of the same scene and then merge them together yourself, or if you let Network start 3 renders on your 3 computers and it auto merges them, the render speed will be the same.

With coop rendering, the Manager sees where the SL of each computer is, and figures out when to stop the renders so that the final merged SL will be what you specified in render options.

The scale between SL is not linear, it takes almost twice as much time to reach the next SL compared to what it took to reach the previous one.
Thanks Mihai,

So Cooperative rendering may actually save time on multiple single renders as the nodes that finish will automatically move on to the next Coop. still render. With more efficient management of node assignment since my nodes are of different processing powers?

Basically I'm not loosing time going with Cooperative rendering over single job rendering is what I'm understanding.
You would probably save time, because like you said if you were to add a job manually of a single scene on your 3 computers (which you later intend to merge renders), lets say you wanted the final merged one to be SL 19. You would have to take a guess to which SL you need to set those jobs so that the final result would be that SL. So maybe SL 17 individually. But maybe that will be too high, and the final merged SL will be maybe 19.7. So you've rendered for too long. Network takes care of that and lands pretty close of the final requested SL, no matter how many computers you do coop with. It won't be exactly 19, but 18.95, or 19.03...something like that, which won't make a big difference if any, visually.
sending rendering to render farm

Hi I work in Rhino 5 and render with Maxwell 4 a[…]


There is a big problem here. I have used Maxell […]

Even with low number of polygons you don't have to[…]

I had some very strong highlights in some material[…]