...
- First steps is to figure out a plan to look at a couple of different options and come back here to decide.
- Cloudbees?
- Building raw resources on AWS or IBM cloud.
- What does this look like?
- The initial idea was an EKS cluster with quotas and namespaces for each team.
- Fabric:
- Interested in a declarative pipeline.
- Long-term: Would like to choose our tools like Concourse, Gitlab-CI based on what we need.
- Short-term: solutions to be discussed and reported on next week.
- Sawtooth:
- Interested in taking existing infrastructure and moving to HL resources provided minimal change is needed to existing pipeline.
- Leave everything at current service provider but move the build minions somewhere else because we suspect they are the problem.
- Moving the minions would solve performance problems.
Action items
- Matthew Sykes Bring a narrowed down list of possibilities to talk about.