Use a selection of use-case oriented policies to apply to your cluster and mix them with intents like rebooting with minimal impact or removing excess nodes.
Run kalc to get exact command sequence that is required to satisfy your current policy and all the previously introduced policies.
Kalc generates YAML diffs in json-patch format that can be applied both in-place to running cluster and to your manifests stored in source control.
You can define your own service level objectives, dependencies and constraints by manipulating state objects using pure Python language, and use the same efficient engine to compute optimal satisfiability scenarios.
Calculate Correct Resource Request and Limits
Automatically Inject Anti Affinity Rules
Estimate cost of your next feature or deployment scaling
Find unused nodes that are safe to switch off