Fix output when deploying from configuration
What does this MR do?
When the Orka provider is configured to use VM configurations, virtual machines don't have names. The name is the name of the configuration, and we get a unique ID after the prepare
step.
So we need to make changes to the way we print this information for it to remain consistent:
- the name of the VM is returned to the runner as part of the
config
stage. This is only used to print out thevia
part of this statement:Running on Gitlabs-iPro via runner-orka-canary-1572770993...
. This MR adds a new provider methodVMNameForRunner
, which is empty in that case, so that the runner only printsRunning on Gitlabs-iPro...
-
VMNameForRunner
returnsVMName
in the GCP provider and in Orka when not deploying from configurations - in the DeployFromConfig case, restore printing the chosen configuration and alias to the build log
- Finally, print the VM ID to the build log in both cases, so finding the VM running a job is easy
before
Running on Gitlabs-iPro via runner-orka-canary-1572770993...
after
Using image macos-11-xcode-12 (aliased to "ork-11-12-202109061639")
Waiting for machine 89323746a3425 to boot...
Running on Gitlabs-iPro...
Why was this MR needed?
What's the best way to test this MR?
From https://gitlab.com/akohlbecker/playground/-/jobs/1589647457
What are the relevant issue numbers?
Edited by Adrien Kohlbecker