This question concerns Maya 2010 and the “5 included mental ray” licenses.

I have 1 network license of Maya 2010 w/ the 5 included mental ray batch licenses. So - I can kick off a mental ray backburner job and while it is rendering away on my 5 render nodes, I can ALSO open the full maya gui on my main workstation, on which I could also interactively render a mental ray scene.

So - I’m in effect using mental ray on all 6 machines.

BUT - if I close maya on the main workstation and start the backburner server (to try to have 6 machines rendering batch) - the main workstation fails to get a mental ray license and fails to render.....???

Huh? That 6th mental ray license I guess if for “gui use only” - but that doesn’t really make sense that I could render in maya but not batch while the 5 other machines are rendering?

Does anyone know a way to be able to have all 6 machines batch rendering a mental ray scene at once?

Many thanks

The fact that this doesn’t work makes perfect sense to me from a technical standpoint, but I do see where it would be nice if it did.  Logically speaking being able to run Maya means you can do everything that batch render is doing, but at the end of the day they are different beasts.

Essentially your network floating license gets you one interactive license + 5 batch render licenses.  Even though from a feature-set point of view one is the superset of the other, you can’t substitute it for the other.

Of course, nothing is preventying you from manually dividing up the frame set and giving the bulk of the work to the Backburner job, and the rest to an interactively kicked of batch render.  Ugly...but I think it would work (haven’t tried).  Obviously that’s not what you are asking for though…

In fact, mayabatch.exe script execution might work as well, in which case you could automate this all from a simple script which starts both “jobs”.

It sure would be ugly though… ;-)

Ryan

Yes I think you are right, too bad there’s not a clean and simple way to do it.  Manually separating frames to render in maya interactive would be a big pain for large renders.

Technically yea it may make sense, but for usability its kinda silly.  I’d be quite happy if using that 6th mental ray license to batch render would then prevent the maya gui from running - that would make sense to me.

Argh, so much for that idea.....

I’ve got 5 machines batch rendering mental ray via backburner.  Opened maya gui on my workstation - although I can render 1 frame, if I try to run a batch render, even though its from the gui, it fails to render “couldn’t get a license.”

So I guess I’m stuck - no matter what I can’t batch render on 6 machines......

Sorry I don’t have an answer for you but if you don’t mind me asking, how did you set up your render farm with maya?  I am in the process of doing the same thing and the documentation is completely confusing on this matter (it is so much easier with 3d max).

Did you just install maya from the DVD on the render nodes?  Did you have to install mental ray standalone on any of the nodes?  The licenses that are included, are they per machine or per processor or per core (I have heard many conflicting stories about this)?  Basically, could you please give me a quick run through of how you set up your farm?

Not sure if this is for David or myself but here’s our setup we are going to try for 2010 (migrating from a working 2009 setup).

We are using a proprietary render / queue manager service under Windows which acts as glue between alienbrain repository and the farm.  Artists can kick off and monitor renders right within alienbrain, and results are automatically imported back in.  The service installs and runs on the repository / render manager server in “master” mode, and this server also is running adskLM network license manager.  This server does not have Maya 2010 installed.

Then on each worker node we have to install Maya 2010 only (no Toxic, Backburner, etc.), along with the same render manager service, this time in “slave” mode.  ICE (Internet Communications Engine) and network shares are used to communicate between the master and slave service instances.  Each slave node runs renders using render.exe.  Each slave node has a .lic file in the C:\flexlm directory that points back to the master server with USE_SERVER.

That’s pretty much it.  We have it set up in parallel with 2009 on the same servers, and both seem to work in parallel.  The 2009 Maya flexLM and 2010 Maya adskLM install side-by-side, and work on different ports by default.

The only problem we are having is when render.exe starts, we are not seeing license pulls on our license server when we run render.exe (which does successfully render).  If we interactively start Maya it does pull a license.  So while everything “works” today; I have concerns that things will stop working after 30 days, as has been reported elsewhere.  If anybody knows if this is normal behavior, or how to fix it, I’d appreciate any tips.

Thanks,
Ryan

The render farm setup varies slightly whether you have a nodelocked maya license, or if you pay the extra $900 or so for a network (floating) license.

If nodelocked - you still run the full maya install on all the render nodes.  You do the install as if you had a license for each one, but you just don’t license them.  You can then network render on them - but maya software renderer only (ie: NOT mental ray.)

If you have a network license - you run the full maya install on the render nodes, and as part of the install you select a server for it to check out a license from.  With a network license you need to install and setup that license server somewhere on your network.  Your license file gets installed on that license server machine.  Your license server machine does not have to be one of the machines you’ll use maya on.

Mental Ray standalone is a separate product, as far as I know, and is not required for just doing straight renders out of maya that use mental ray.

The “5 included” mental ray licenses (only if you have a network license) are per machine - regardless of how many cpu’s or cores it has.  (I believe part of the confusion here is that wasn’t always the case.) So with 1 network license of maya 2010 you can have 5 machines batch rendering with mental ray, *and* maya gui open and in use on 1 machine at the same time.

Ryan - Its interesting that renders started with Render.exe are not checking out a license.  But yes, I suspect after 30 days that will stop working.  When I first got maya 2010 it was still nodelocked but was surprised it all worked at first, 30 days later I could no longer network batch render mental ray and had to pay to upgrade to a network license.

And to make matters worse I found out with Max you can network batch render MR on an unlimited number of machines......  Why such different licensing policies for two 3D apps (now) made by the same company?

I know its not.

But Autodesk licenses Mental Ray both for 3DS Max and Maya.  Why such different licensing setups and pricing between Max and Maya?

Buy a seat of Max and you get unlimited MR network batch licensing.

Buy a seat of Maya, pay $900 more for a network license, and get only 5 MR batch licenses.

We bought some Maya 2010 licenses. One of them is an Network License.

Besides that we’ve installed the 5 batch nodes on renderfarm machines.

When i use backburner even if on the monitor i see more than five machines i can only use 5 of them (no mather if they’re SLM or batch nodes).

My question is this. Is there any way that i can “upgrade” my license so that i can render with more than 5 machines? It’s terrible because i want to use all my Maya’s and i’m limited to 5 machines, either Maya SLM or batch nodes.

If i use renderpal or any other Render Management Software i can render with my NLM and all Maya SLM but it doesn’t recognize my 5 batch nodes.

I just want to be able to use all my licenses and not being limited by this backburner issue.

From my point of view i should be able to use both my Maya SLM licenses and my batch nodes that came with my Maya NLM.

Can someone please help me out on this?