We have an on-machine HMI that runs the Ignition 8.0.2 gateway as well as a single client on a limited 1 client system (OS is Windows 10). A full screen app shortcut created by 8.0.2 Vision Client Launcher on desktop starts the HMI as expected. We copied this shortcut to C:\Users\Administrator\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup, expecting it to launch the client on Windows startup as on other units running Ignition 7.9.9. Instead, we’re getting this (which we don’t see when we run the shortcut ourselves):
Here’s the app config with names mostly blanked out (we tried increasing Timeout, but it doesn’t seem to apply at the stage this is failing–tried 600 s and still the same result, though Windows restarts in less than half that time):
We can run a cmd script on startup to run the app after a delay, but I expect this isn’t supposed to be necessary (see this post–though I don’t think it’s the same kind of problem this time as it occurs whether or not there was a client running immediately before restart). Is there something we can adjust/change to get this working, or is this a bug/regression in the 8.0.2 Vision Client Launcher?
FYI we get the same error message if we try to manually start the app too soon after Windows restart. It’s ready somewhere between 30 and 60 seconds after Windows startup and we no longer get the error message if we wait long enough to start it (via script or manually). It’s definitely an issue of client launcher not waiting for gateway to reach running state as the following script resolves the issue:
:: Checks for Ignition Gateway running before opening app.
:: Run from link in "C:\Users\userName\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup"
set clientLauncher=C:\Users\Administrator\AppData\Roaming\Inductive Automation\Vision Client Launcher\visionclientlauncher.exe
set app=App Name
set /a count=1
echo Waiting for Ignition Gateway...
powershell -Command "(new-object net.webclient).DownloadString('http://%host%:%port%/main/StatusPing')" | find "RUNNING" > nul
if errorlevel 1 (
set /a count+=1
echo Check %count% - Ignition Gateway is starting...
timeout 1 > nul
) else (
echo Ignition Gateway is running. HMI starting now...
:: Leave message on screen long enough for operators to see app starting.
timeout 10 > nul
It looks like this regression should be resolved in 8.0.6, though I won’t be testing till 8.0.6 is the stable release:
15036 Launcher respects retries, added new modal with feedback Fixed an issue where Client Launchers weren’t respecting timeout and retries settings when attempting to target a gateway that isn’t running. Additionally, added a new modal that provides feedback about the attempts
I’m glad the batch file got it up and running, but am surprised you’re seeing this on 8.0.7. We’re also running 8.0.7 and have not seen this issue since 8.0.5. Did you try setting the timeout higher on the application link (screen clipping I shared above just sets defaults for new application links)?