You need to log in to create posts and topics.

3dsmax, jobs won't reach coordinator


I'm actually testing Pandora on our farm (14 pcs on windows10 pro) and i'm facing some troubles.

The farm is functional, I installed pandora on every machines with only one with the coordinator. In the render handler I can see all the machines in idle state, so no problems. Troubles comes as soon as I try to send a job on the farm from 3dsmax;

Firstly, in the "Outputpath" Pandora asks for a complete filename, not only a folder structure, but my file type is already set inside 3dsmax, so I don't want to define a new one (additionally, Pandora doesn't have any option regarding image file format (that I know of)). But to test I put something like that (on a network place accessible by every machines) "networkmachine\render\testpandora\testpandora.exr" I put a name in "projectname" and another one in "jobname".

When submitting the job, it seems like it does something (in fact it does as it copies the .max file in the network drive workstations/nameofthesubmittingmachine/Jobsubmissions/fftapjtjpl/) but nothing else happens, nothing in the renderHandler, no files created, no max started on the slaves..........does not work.


I got nothing on the coordinator debug log.

I'm obviously doing something wrong here but I'm quite stuck.

Ok I still don't know how to use the output name from max (and avoiding being forced to define a new one in "Outputpath") but I find out why the renders did not launch:

I got an accentuation on the .max file name, and Pandora does not like that at all. Accent removed, the rendering started, good.

Hence I got another trouble but found the workaround: some machines on the farm (the render blades) are not connected to the internet, only on the local network and those machines refused to take the job...... strange...

But after some thoughts I noticed that the time on those machines was not the same as the ones connected to the internet (because, obviously they take the time infomations from an online reference), so I just synchronized the non internet connected machines to one of the workstations (witch is connected) and voilà!!

Here's the cmd (with administrator privileges): NET TIME \\nameofthereferencemachine /SET (don't forget to set the correct time zone on date/time preferences)


I'm happy to hear that you could already fix the issues. The jobassignments expire after 15 min. If the system time is different on the slaves they think that the task was assigned hours ago and so they refuse it and wait for a new assignment.

At the moment there is no option to read the outputpath from the rendersettings, but I'll add that probably in the next update.

Thanks for the cmd to fix the time issue.