Would love to see the following scenarios explored too: - failure of a backend node - adding additional capacity From experience both scenarios can trigger n

Would love to see the following scenarios explored too: - failure of a backend node - adding additional capacity From experience both scenarios can trigger non-obvious behaviours (especially when you have backends with different power levels).