I have an AWS auto-scaler group using docker-autoscaler executor. I have the desired capacity set to 1, min set to 0 and max set to 15. For some reason my gitlab jobs are not executing concurrently on the same instance. Once the EC2 instance completes running the job, the instance terminates and the desired capacity is decremented to 0 (the min). The remaining jobs in my gitlab pipeline are left waiting for an instance until it times out. I also receive an error on my gitlab-runner.service that shows the following:
builds=0 error=failed to update executor: reserving taskscaler capacity: no capacity: no immediately available capacity executor=docker-autoscaler
In order to get the other jobs to run, I have to manually update the desired capacity once again. I've also tried setting the min to 1 and the desired to 1, but it still terminates my instance and gives me an error on the gitlab runner that it tried to decrement the desired capacity but it couldn't because it is as the minimum value.