Spot instance interruption handling

Make Spot instances viable for more workloads by handling interruptions gracefully — the more workloads you can safely run on Spot, the greater your savings.

For a sample pod configuration with a graceful termination grace period, see this spot-interruption-handling.yaml , a preStop hook to drain connections, and topology spread constraints for resilience across Spot pools.

For the complete scripts and manifests, see the 10-additional-cost-saving-strategies folder in the code repository.