ADR002: Concourse pool resource
Context
When building pipelines using concourse, we investigated using the pool resource in order to control flow through jobs. This was an alternative to the use of the semver resource.
These 2 resources are both workarounds to solve the problem of triggering jobs when we haven’t made changes to a resource.
The problem is that the pool resource relies on write access to a github repo, which means we must pass public keys that allow this access into the pipeline and deployed concourse instance - we want to minimise the number of credentials we pass, and the semver resource relies on AWS credentials that are already passed.
Decision
We will not use the pool resource for flow between jobs - instead we will use the semver resource
Status
Accepted
Consequences
This was an investigation into a different approach, so no consequences