I have a repo set up locally to track a remote repo on AWS Code Commit. This repo has two permanent branches: 'master' and 'development'. I am trying to deploy each branch to separate Elastic Beanstalk applications, where one is used for production and the other for testing.
For each branch, I've used the EB CLI tool to create an application instance that tracks the appropriate remote branch in code commit. The first branch I do this with (master) always works, but once I add an application to track the 'development' branch I seem to create a duplicate remote called 'codecommit-origin', which I do not have access to edit. Example:
codecommit-origin https://git-codecommit.us-region-1.amazonaws.com/v1/repos/some_repo (fetch)
codecommit-origin https://git-codecommit.us-region-1.amazonaws.com/v1/repos/some_repo (push)
origin ssh://git-codecommit.us-region-1.amazonaws.com/v1/repos/some_repo (fetch)
origin ssh://git-codecommit.us-region-1.amazonaws.com/v1/repos/some_repo (push)
I do not have permission to access the 'codecommit-origin' remote, and removing it removes my ability to deploy the application, throwing this error:
ERROR: AttributeError :: 'NoneType' object has no attribute 'split'
For obvious reasons, this makes deploying updates to my application pretty painful. I essentially have to reset my remote with "git remote add origin [url]", push my commits, run "eb init" again, deploy, then remove the "codecommit-origin" it creates every time I want to change something.
Is there a better way to manage this workflow of deploying two elastic beanstalk applications from two branches in a single code commit repo?