Description of problem: Right now the only possible way to set an environment variable is through a pre-hook, which requires the variables to be in git. If I can put them into git, I can put them into my code directly. The beauty of ENV variables is a) I don't need a new commit just to change a variable b) They are protected. I might have 20 developers of which NONE are allowed to know our database passwords and Facebook secret ids. git is NOT a solution to this kind of data. Overall, git is not built to protect sensible information. Version-Release number of selected component (if applicable): latest as of today How reproducible: See https://openshift.redhat.com/community/blogs/new-openshift-release-may-15-2012-community-code-custom-environment-variables#comment-24214 for current recommendation and comments for flaws. https://openshift.redhat.com/community/content/custom-environment-variables#comment-24215 Idea / request Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
We've got a story to cover this, and it's being tracked internally. We're working on getting the roadmap elements laid out online. Transferring to Krishna to organize and get slotted into the roadmap.
Tracking in US1821