Open fnmarquez opened 2 years ago
There is no auto-scaling feature at this time.
Something you can do is patch the deployment as needed. Depending on your needs you could do something like scale it up or down during certain times of day (using a cron workflow) or maybe at the beginning/end of a workflow run.
You can patch the replicas count as follows:
$ oc patch deployment actions-runner-1627686781 -p '{ "spec": { "replicas": 2 } }'
deployment.apps/actions-runner-1627686781 patched
$ oc get deploy
NAME READY UP-TO-DATE AVAILABLE AGE
actions-runner-1627686781 1/2 2 1 3m37s
The deployment name is the same as the Helm release name, or you can do:
oc get deploy -l=app.kubernetes.io/instance=$RELEASE_NAME
Is it possible to autoscale runners on demand? (not manually from the workflow)