Closed iper-jkane closed 3 months ago
main changes are: putting keycloak into separate folder / Docker config; whilst still allowing overrides via env vars allowing the use of an env var to select the base image (potentially scriptable based on architecture) switching to a more recent version of the official dockerhub image generating it's own test certificates but allowing keycloak to run in production mode points the ui at the local keycloak for non LS login returns results and filtering terms from most endpoints hopefully there's something useful in there ;)
oh. sorry. I've just noticed! I didn't really want to try and push to EGA-archive:master wanted to push as separate branch. :O :D
Basically, I was having trouble getting the web interface to talk nicely to keycloak + permissions, this was my attempt at solving locally, and was hoping it might be useful "upstream".
oh. sorry. I've just noticed! I didn't really want to try and push to EGA-archive:master wanted to push as separate branch. :O :D
Basically, I was having trouble getting the web interface to talk nicely to keycloak + permissions, this was my attempt at solving locally, and was hoping it might be useful "upstream".
Hi @iper-jkane ! Thank you for the PR. I see this is mostly about configuring a new and different deployment, some parts of which I see it's useful. Nevertheless, I can't merge it all to the master branch. Do you want me to create another branch (so you can deploy and continue developing from there) and then maybe I can cherry pick what I see useful for the RI? Best, Oriol
changes made to test local keycloak + frontend installation basically working hostname of test/dev server is 'beaker'. which can be sed/replaced fairly easily. e.g., to see files that require config changes try, (from root of beacon2-ri-api folder):
> grep --exclude-dir .git beaker . -r