Pipelines executed on Docker Machine executor occassionally fail with `no such remote 'origin'`
Zendesk: https://gitlab.zendesk.com/agent/tickets/103864
A customer is using Docker Machine executor on AWS EC2 and the following error occurs regularly. If they kill the given EC2 instance and let the Runner manager rebuild on next pipeline run the error goes away for a time. The error no such remote 'origin'
seems very odd. What would cause that?
Running with gitlab-runner 11.1.0 (081978aa)
on gitlab-runner-manager 53adede2
Using Docker executor with image python:3.6 ...
Pulling docker image python:3.6 ...
Using docker image sha256:4f13b7f2138e95e8f46086463924252ccf24621c60123475970d06c06dfdcba7 for python:3.6 ...
Running on runner-53adede2-project-93-concurrent-0 via runner-53adede2-gitlab-docker-machine-1536974965-b16550b3...
Fetching changes...
HEAD is now at 8aa515c added pw info
fatal: No such remote 'origin'
ERROR: Job failed: exit code 1