Scenarios¶
Real-world hibernation setups, ordered from the simplest to the most advanced. Each scenario is a self-contained page with complete, copy-pasteable YAML — connectors, plans, exceptions, and notifications — framed around a concrete business situation.
Assumptions: you have the Hibernator operator installed and understand the basics from the Quick Start.
The Scenario Ladder¶
| Tier | Scenario | What it demonstrates | Link |
|---|---|---|---|
| Basic | Nightly Full Shutdown & Wakeup | One plan stops an entire environment every night and restores it every morning | Details |
| Basic | Weekend Hibernation | Keep a dev environment off from Friday evening to Monday morning | Details |
| Intermediate | Partial Hibernation | Keep critical services running while everything else hibernates | Details |
| Intermediate | Tolerating Partial Failures | Let a large fleet hibernate even when some targets fail | Details |
| Intermediate | Suspend Scheduled Shutdown | Temporarily suppress a nightly window for testing or incidents | Details |
| Intermediate | Selective RDS Hibernation | Hibernate only non-critical dev/staging databases with tagSelector | Details |
| Advanced | Dependency-Ordered Shutdown (DAG) | Databases stop last and wake first via explicit dependencies | Details |
| Advanced | Staged Hibernation by Criticality | Roll hibernation through tiers: dev → batch → shared | Details |
| Advanced | Event-Mode Overrides | Change which targets hibernate (and how) during a special event | Details |
| Advanced | Weekend Subset Wakeup | Wake only 2 of 10 databases during a suspend window, auto-revert after | Details |
| Advanced | Holiday Schedule Replacement | Replace the entire schedule and execution strategy for a period | Details |
| Expert | Dry-Run Rehearsal | Validate a production-shaped plan with zero cloud impact | Details |
| Expert | Production-Grade Setup | Cross-account access, ordered shutdown, Slack alerts, governance | Details |
Capability Matrix¶
Use this matrix to find the scenario that exercises a specific feature combination.
| Scenario | Strategy | Behavior | Exception | Executors | Key feature |
|---|---|---|---|---|---|
| Nightly Full Shutdown | Parallel | Strict | — | eks, rds, ec2 | Full lifecycle |
| Weekend Hibernation | Parallel | Strict | — | eks, rds | Schedule windows |
| Partial Hibernation | Parallel | Strict | — | workloadscaler, karpenter, eks | Selector granularity |
| Tolerating Partial Failures | Parallel | BestEffort | — | workloadscaler, ec2 | Failure tolerance |
| Suspend Scheduled Shutdown | Parallel | Strict | suspend | eks | Carve-out suppression |
| Selective RDS Hibernation | Parallel | Strict | — | rds | Expression-based selection |
| Dependency-Ordered Shutdown | DAG | Strict | — | workloadscaler, eks, ec2, rds | Ordering |
| Staged by Criticality | Staged | Strict | — | workloadscaler, karpenter, eks, rds | Stage groups |
| Event-Mode Overrides | Parallel (base) | Strict | extend + targetOverrides | ec2, eks, rds | Per-target overrides |
| Weekend Subset Wakeup | Parallel (base) | Strict | suspend + targetOverrides | rds | Partial wakeup subset |
| Holiday Replacement | Sequential (override) | BestEffort (override) | replace + executionOverride | eks, rds | Schedule + strategy replacement |
| Dry-Run Rehearsal | DAG | BestEffort | — | noop | Failure simulation |
| Production-Grade Setup | DAG | Strict | replace (referenced) | workloadscaler, karpenter, eks, rds | Cross-account + notifications |
How to Use These Scenarios¶
- Copy and adapt. Every YAML block is complete and internally consistent. Account IDs, regions, cluster names, tags, and labels are illustrative — replace them with your own.
- Selectors are the contract. Most scenarios select resources by tags or labels (
Environment,workload,critical, ...). Adjust the selector shapes to your tagging/labeling conventions, keeping the structure intact. - Start simple, climb the ladder. Each tier builds on concepts from the previous one. If you are new to Hibernator, start with the Basic tier even if your end goal is an Expert scenario.
- Concepts are linked, not repeated. Scenario pages explain the situation and the setup; they link to Concepts and User Guides for deep dives instead of duplicating them.
GCP executors
The gke and cloudsql executors are still pending API integration and are not covered by any scenario. See the roadmap for their status.