Utilizing the Vision Client with a Remote Desktop Server

Looking at trying to get away from putting full PC's out on the shop floor and replacing with something that requires less maintenance. Set up a test environment using linux and xrdp to create a remote desktop server with the Pi device that powers up and connects to. Works really well, the issue we are having is that we have a couple projects that everyone runs in the Vision client and they are connected to the work cell based on the host name. When trying to do this from the remote desktop server, all the clients are showing as from the same host name and we are getting the error that the project is already running. Worked a bit with support on this and not having any luck.

Anyone have any luck with anything similar that worked for them?

From inside the client (indicating you're running a limited Vision license) or from your RDP software?

Either way I'd recommend contacting our sales engineering department or doing some digging here on the forums in case someone's already documented a solution.

It is not a limited license. I am forwarding our support case on to our sales engineering contact this afternoon. Just wanted to look through forums to see if maybe someone had overcome this issue.

1 Like

Is the gateway running on the Pi? Or is everyone RDP-ing into the Pi to get to the gateway?

You throw away a bunch of useful functionality when you run Vision in an RDP server--particularly the physical unique location detection it can do. Instead of an RPi, consider using a modestly-equipped fanless X86 brick PC. (RPis can run Vision directly, but you have to use a custom launcher.)

Using an RDP server means you need Windows on your plant's production network, which is engineering malpractice, IMNSHO. Ok, not if using xrdp on linux. But still not great.

2 Likes

Everyone is using xrdp on the pi to access a desktop session on the Ubuntu server

Any info on this "custom launcher" would be greatly appreciated. We have tried several iterations of trying to get it to run on the pi itself. The issue we always ran into is that we use the web module in ignition to assist with translations and it always fails when that loads.

Yeah, sorry. That won't run on ARM.

If you want to get deeper into the weeds, you could look into VDI, where each client RPi would be assigned its own VM instance running on the Ubuntu server and would therefore have a unique hostname.

We're running a remote desktop server on Windows 2022 with multiple clients just fine. It may be related to using Linux instead of Windows.

It sounds like you are doing a hostname check in the client startup script and rejecting duplicates based off of that.

You can always modify that and allow multiple connections from the RDP host name.

Is that a setting in Ignition?

In Designer for the project you are seeing this happen on, go to Vision-Client Events-Startup.
That will be the most likely location for a host name check.

3 Likes

So as it turns out there was an entry in the project that was looking for hostname's that was causing the issue. Thanks for the all the help on this!

1 Like