We should make runners on us-west-1 dynamic:
When there is a substantial queue for a runner type (20m?) the scaleset for that runner type should go from max_runners 0 -> unlimited, no-flap mechanism should keep it for at least 3h live with low queue before returning to max_runners 0;
We should make runners on us-west-1 dynamic:
When there is a substantial queue for a runner type (20m?) the scaleset for that runner type should go from max_runners 0 -> unlimited, no-flap mechanism should keep it for at least 3h live with low queue before returning to max_runners 0;