I’m trying to set up a job (through commandline) that that will precalc a vray irradiance map. Everything is going exactly as it should…except no vrmap file is saved out at the end.
I’ve included these in my info file. There are no errors in the job.
vray_options_dontRenderImage=true
vray_adv_irradmap_mode_enums=“Incremental Add to Current Map”
vray_adv_irradmap_dontDelete=true
vray_adv_irradmap_autoSave=true
vray_adv_irradmap_autoSaveFileName=“C:\example.vrmap”
Any help?
Jeff
Hey Jeff,
We can definitely look into this for you and give you a hand in figuring it out 
Could you send us the Job/Plugin Info files you are using for your submission, in addition to the slave log from the slave that picked up the job, so that we can get a headstart on this one? Thanks in advance!
Cheers,
PM’d
Hey Jeff,
Got the files you sent, am looking at them now. Something I’ve noticed is that our plugin doesn’t grab and set the adv_irradmap_autoSaveFileName setting for vray – though this line wasn’t in the info file you sent me (only in your original post here), so I’m guessing you figured that out
We’re also not setting the adv_irradmap_mode_enums, and that one is still in the info file. I’m not familiar enough with vray/Max to know exactly what these options do, or why we excluded them from our plugin. I’m trying to figure out the latter now, but I figured I’d post this to let you know of my progress, and maybe give you a clue as to what could be the problem
Cheers,
the adv_irradmap_mode_enums setting is which irradiance mode vray should use. This has to be set to 4 -“incremental add to current map.” This means that instead of starting a new irradiance map for each frame, it will combine it with the existing one in memory from the previous frames.
When it didn’t look like deadline was going to be able to do it with the existing settings, I decided to attempt setting all the parameters manually through maxscript. That file is also included. But still it fails to save it. I’m almost wondering if the way deadline opens and renders scenes it prevents vray from doing what it needs to do, because it seems to be a bit different than the normal way that backburner does it.
Hey Jeff,
Sorry, didn’t notice your custom maxscript in there! I’m guessing you’re running it as a PostLoad script for the job? If not, I would recommend you trying that – you can do this by addding the ‘PostLoadScript’ key to the plugin info file with the name of the script file as the parameter. In addition to this, you would also need to submit the script with the job by appending it to your submission command. This will execute the script after the scene file has been loaded, but before it renders it.
One thing I noticed in your custom maxscript file is that you’re using single backslashes for the pathnames – I’m pretty sure Max uses backslashes as an escape character, so you’d either want to use forward slashes (ie, ‘/’) or double backslashes (ie, ‘\’) for your paths. So they would look something like:
\\\\a64\\ClientFiles\\NewUser\\Scenes\\VrayIRR_Test\\VrayIRRTest26.vrmapor
//a64/ClientFiles/NewUser/Scenes/VrayIRR_Test/VrayIRRTest26.vrmapNOTE: I’m not totally sure it’s a quadruple backslash for the network paths – if it doesn’t work, try only 3.
Hopefully that solves the problem – if not, I’m going to have to keep digging 
Cheers,
I will definitely look into this. I also just realized that vray creates its own log file. It never occured to me to check to see if there were any errors in that file. I’ll let you know what I find.
I’m glad you mentioned that. after looking at the pathing issues and determining that wasn’t it. I was looking back over the logs and didn’t notice a line say running maxscript file or whatever it says. Sure enough I had written that file and never actually put it into the info, so it was never being used…silly me. Put it in, instant workage. thanks for the pointer much appreciated.
Okay now that I’ve got it working correctly had another question. One thing that is vital for doing irradiance maps is it must be computed on one machine… uninterrupted. Because its using the existing irr map in memory from the previous frames. I’ve set the machine limit to one, and that takes care of it most of the time. But on occasion the machine rendering it gets taken offline by mistake or crashes. Is there a way to tell deadline that if a job gets interrupted and needs to be sent to another machine…it starts the job over from the beginning?
Jeff
Hey Jeff,
There isn’t this kind of functionality inherently built into Deadline, but what you could do is set the “Frames Per Task” option in the submission script to however many frames you are rendering. This should ensure that there’s only one task in the job, so if the slave currently rendering it crashes for whatever reason, a different slave will pick up the job and have to start the one task from scratch.
Hope this helps!
Cheers,
Is that the same as the chunksize setting? Cause I thought of that and tried setting that once and it didn’t work when rendering in nth frames. When doing an irradiance map you render out every 5th or 10th frame usually, and deadline doesn’t seem to be able to do chunks when you render that way.
Hey Jeff,
ChunkSize is indeed the same thing – it’s generally labelled “Frames per task” in our submission scripts, I forgot it was called “ChunkSize” in the info files
I just tested this on our end, and you are correct – the ChunkSize setting does not work when specifying individual frames (eg, “0, 5, 10”) as opposed to a range (eg, “0-10”). I will have to look into whether this is a bug, or if it was excluded intentionally. Either way though, I guess there’s currently no way (at least until we fix this) for you to force the whole job to be restarted if the slave rendering it crashes/restarts if you’re specifying individual frames 
Cheers,