Attune is a Kubernetes operator that right-sizes pod CPU and memory without restarts, using In-Place Pod Resize (Kubernetes 1.33+ beta, 1.32 alpha). It replaces the Vertical Pod Autoscaler with something that actually works in production.
| VPA | Attune | |
|---|---|---|
| Resize method | Evicts pods | In-place (no restarts) |
| HPA compatible | No (death spirals) | Yes (adjusts base, not %) |
| Safety | Minimal guardrails | Auto-revert on OOM, throttle, restarts |
| Confidence | Backward-looking histograms | Time-of-day-aware + graduated rollout |
- Zero-downtime resizing via the
/resizesubresource - Composable recommendation engine with percentile, confidence, margins, and change filters
- HPA-aware conflict detection prevents horizontal/vertical scaling fights
- Safety monitor with automatic rollback on OOM kills, CPU throttle, or restart spikes
- Graduated rollout: Observe, Recommend, OneShot, Canary, Auto
- Multi-source metrics from Prometheus, Thanos, VictoriaMetrics, and Grafana Mimir
- Cost savings estimation per workload with
kubectl attune savings - Helm chart with cert-manager webhook support and Grafana dashboard
helm install attune oci://ghcr.io/attune-io/charts/attune \
--namespace attune-system --create-namespace- Documentation -- full guides, API reference, and examples
- GitHub Repository -- source, issues, and contributing
- Migrating from VPA -- step-by-step replacement guide