We were using Jenkins as our CICD pipeline. In Jenkins the executions are independent of each other and dont supersede each other. This gave us the flexibility to deploy to production the changeset that has been tested.
I am trying to migrate to AWS codepipeline and cannot get my head around this issue now i.e. how to make every execution independent.
my pipeline structure looks something like this:-
Stage1 - Download source from Git
Stage2 - Test and Package
Stage3 - Approve deployment and Deploy to Test
Stage4 - Approve deployment and Deploy to Prod.
Now if I have three changeset (Changeset 1, Changeset 2 and Changeset 3 - with Changeset 1 being earliest and Changeset 3 being latest). If I have a situation where I need to deploy changeset 2 to Production(As Changeset 3 is not tested fully) , then this is not possible as Changeset 3 would have superseded Changeset 2.
Please let me know if you have any suggestions to overcome this as our team would like to perform continuous deployments to prod for tested changes. Is there a need to restructure the pipeline which can resolve this issue?
AWS OFFICIALUpdated 2 years ago
