We can add various rules in the autoscaler:
- The trigger for scaling: do we want to base scaling decisions on CPU usage or on the length of a given queue?
- The scale up and scale down rules: do we scale by one instance or add / remove 5 at a time?
- The interval: how long do we want to not touch the number of instances running after the previous scale operation?
- The range: what’s the minimum and maximum required instances we want to have running?
A long awaited feature is there! I'll enable this for some services and see how it goes...
This is an imported post. It was imported from my old blog using an automated tool and may contain formatting errors and/or broken images.