Does anyone have experience in running head-less jArchi script as part of an Azure DevOps pipeline ?
I've built a Docker container image with a fresh copy of Archi & jArchi in head-less configuration and call it inside scripts to consolidate many Archi views into one. This work perfectly well under Ubuntu on my station, also under Windows WSL2 context.
The same code hangs in Azure DevOps, where the script goes up to the first "docker run... archi" statement and never exits it... it runs forever... Other non-Archi containers images also run normally in the pipeline.
Thanks for your ideas!
Found how to manage this.
When following the rules to create a customized Archi package (see How to create a packaged version of Archi (https://github.com/archimatetool/archi/wiki/How-to-create-a-packaged-version-of-Archi-(including-configuration-and-plugins))), the 'custom_defaults.prefs' file should be located in the current directory where you actually start Archi for the first time and not in the directory where Archi is installed.
- a sample of azure-pipelines.yml configuration file, where the 'archimatetool' directory contains the installation files (Archi-Linux*.tgz distribution, Archi.ini file , dropins folder, custom_defaults.prefs file)
- a test bash scripts, located in a 'scripts' directory, that invokes Archi and a test.ajs script (not provided - it does you want)
- any jArchi scripts, located in a 'jArchi' directory