Goal

• Identify a set of steps and actions necessary to get rid of the most painful maintenance burdens of recert, especially for OCP component teams
• Identify change set that can be transferred to RFEs that is able to give recert a more "sustainable" future, e.g. by defining mandatory tests, reducing time to break/fix the recert behavior and avoiding regressions when changing flows that affect recert

Why is this important?

• We need to enable all OCP component teams to manage their own interaction with recert without domain knowledge of it.
• EE will not be able to continously remediate all regressions so its important that we change recert into a software that can be tested individually by component teams without major overhead.
• To scale out the maintenance of individual steps for recert, we need to reduce recerts change scope in the etcd to a well defined set of resources and files that are its "API". After this we can use it for regression testing.

Scenarios

1. Have any central OCP component team regress on recert, and want to fix a pipeline error. For this one needs
   1. a pipeline to regress on
   2. a possibility to detect regression on your own component safely and easily
   3. a way to easily test (break/fix) and verify the regression in CI and on AWS clusters.

Acceptance Criteria

• All CIs for component teams that must be written are identified
• All Epics for changing recert and the individual components are created
• Acceptance from the IBU team on a POC that showcases one of the component changes and how recert can then be tested.

Dependencies (internal and external)

N/A

Previous Work (Optional):

1. Recert Entity Overview
2. Overview on Problems & Solutions on Sustainable Recert

Open questions:

N/A

Done Checklist

• CI - CI is running, tests epics are created.
• DEV - POC created and changeset shared: <link to meaningful PR or GitHub Repo>
• DEV - Approach signed off by staff engineering and Telco Engineering