You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, the EGO public key is configured via env variable. This means, the person deploying dms-ui needs to:
call the ego public key endpoint
extract the public key from the response
update the dms-ui qa and staging helm values
make a PR
get someone to review the PR
merge pr
redeploy the new public key
It would be better if instead dms-ui required EGOs public key url, and then on boot, it fetches the public key and uses that. If the public key changes, instead of going through the complicated steps above, the dms-ui just needs to do a rolling restarts (i.e k rollout restart deploy/dms-ui). This significantly simplifies the deployment process
The text was updated successfully, but these errors were encountered:
After discussing with @anncatton and cnosultation with @hlminh2000 , it seems the EGO public key is actually currently NOT used at all in the DMS-UI code. The config is in the helm chart but is not required for any functionality. In the future, if we implement access control, this may be required, but for now it leads to confusion during deployment. Suggestion is to remove it from the DMS-UI code and the Helm chart config.
At minimum, we'll backlog it and discuss with @rtisma, but this will not be included in the next sprint.
Currently, the EGO public key is configured via env variable. This means, the person deploying dms-ui needs to:
It would be better if instead dms-ui required EGOs public key url, and then on boot, it fetches the public key and uses that. If the public key changes, instead of going through the complicated steps above, the dms-ui just needs to do a rolling restarts (i.e
k rollout restart deploy/dms-ui
). This significantly simplifies the deployment processThe text was updated successfully, but these errors were encountered: