Large Scale MilluNode Deployment Questions / Issues

Hi,

I'm a video playback operator who works mainly in film and television, and I was exploring the possibility deploying Millumin along with MilluNode to push graphics (both images and videos) to several monitors on a film set. Think for example of an office set with 100+ computer monitors, each with a Mac Mini attached to them outputting a mixture of still and motion graphics for background actors to "interact" with / look like they're working.

I understand that MilluNode is still in beta, but wanted to share my findings on some of the weird setup quirks as well as some issues I found along the way in trying to implement this. Not entirely sure if these are bugs, or maybe I'm just pushing the software beyond its intended use. But I'm open to any and all suggestions, especially if I'm going about this in a suboptimal way. All that said, I do see huge potential for using Millumin for this type of use case if some of these things could be worked out.

Some things to note:

• All displays / monitors are connected to Mac Minis.
• Mac minis are hardwired to the same LAN, all running MilluNode.

Here are some of the setup quirks I rant into:

• On each computer running MilluNode, you have to rename the machine alias under Preferences on MilluNode to a unique name, otherwise the Millumin project will show several "MilluNode" instances when selecting an output for each canvas. What I think would be easiest and most intuitive is to have MilluNode use the computer's name as the machine alias. If you wish to change it to something else, then that should be an option, but defaulting to the computer's name would save a lot of time when doing this in any scale.
• On the Millumin project, when selecting the output for the canvas, you have to manually type in the canvas resolution, and by default the display for the output is set to "no display". This adds up to a lot of time when having to manually input the resolution for each canvas and selecting the display for each output. It would be nice if after selecting a MilluNode computer from the output list that if there are any available displays, it is automatically selected, and for the resolution to automatically match the display. Or at the very least, having a general setting to change the default resolution of the canvas away from 720p.
• Additionally, and I'm aware this has been previously brought up by other users in this forum, each canvas can only be routed to a single MilluNode machine as an output. Ideally there would be a way to route one canvas to multiple MilluNode machines, as the same graphics could be randomly assigned across different monitors, especially in a big set where it is all out of focus and in the background of the camera shot.

Here are some of the issues I ran into:

• I initially set up one single Millumin project where I could see all of the different canvases / computers running MilluNode, and noticed that when relaunching the project, most if not all MilluNode computers stall when I try to change columns in the project. It usually took 10-15 minutes before response would become snappy again and all computers running MilluNode would change instantly when switching columns.
• In attempting to troubleshoot the issue, I would reboot the computers running MilluNode, only to realize that after they rebooted and MilluNode would run, the original canvases that were using those computers as outputs would unlink, and I would have to reselect them as outputs every time. When running several computers with MilluNode, this becomes a huge pain point and a lot of wasted time. I'm wondering if it's possible for the canvases to relink to the outputs they were previously paired to as soon as they become available.
• I tried segmenting the original project into a few separate project files to see if that would improve the performance, and did notice that the less canvases / MilluNode computers running off of a single project, the quicker all of the MilluNode computers would become fully responsive after reopening the project and changing columns. That said, it seems like anything over 10 canvases seems to stall for at least a few minutes after reopening the project. I'm curious if this issue is due to something with the LAN, the main computer running Millumin, the several Mac Minis running MilluNode, a combination of the above, or something else entirely?
• It also seems like every time computers running MilluNode restart, the media from the project has to reloaded on the MilluNode machine. Is this the case? Is there a way to have the media permanently copied to the MilluNode machine so that it doesn't have to request media from the Project?
• Lastly, every once in a while, when switching projects and/or switching columns, some of the MilluNode computers would switch out of full screen mode and into windowed mode. Pressing Command + F on the project wouldn't change anything. Only thing that would fix it is quitting MilluNode and reopening it.

Thanks in advance for taking the time to read through this!

Zé