Approaches for implementing and managing patterns for Bitbucket Pipeline projects

[Michael Villis]

Hey Everyone. I'm searching for insights from other Bitbucket Pipeline users on how they manage pipelines at scale across all their teams.

As part of our rollout we are thinking of defining a series of templates for our different app types (eg. Angular website, Spring Boot API) and each will have a code repository skeleton and also a pipeline skeleton. Our developer engineering portal can trigger these to be rolled out as needed.

This all seems pretty straightforward. Some of the things that I'm curious about are:

• How to manage pipeline.yml drift as either teams customize their pipeline or new versions of the pipeline skeleton are released. I could see something like renovatebot for pipelines being useful.
• How to ensure important steps like change control approvals aren't removed from the pipeline? Maybe a combination of pull request controls and static analysis rules?

Any thoughts appreciated.

[Michael Villis]

It looks like part of Atlassian's solution for this is this upcoming feature:

https://www.atlassian.com/wac/roadmap/cloud/reusable-ci-workflows?selectedProduct=bitbucket&p=ca8f751c-aa 

Looking forward to more details