Run entire pipeline by runners of the same physical machine
Description
Our pipeline is split into multiple stages, some stages have the same tag, some not. We have runners on two physical machines listening for jobs of these stages.
Unfortunately, we wrongly assumed that a pipeline's jobs would be executed by runners of the same physical machine. Thus, some dependency stages are being executed on machine A, while jobs that depend on these stages run on machine B. Our pipelines are failing because of missing files/artifacts/dependencies..
Proposal
Add a pipeline configuration to run all jobs on the same physical machine once pipeline has started
Links to related issues and merge requests / references
Edited by Paul Addy