Deployment in your environment
Choose Helm or Compose, connect your database, storage and registry, then check health, backup, recovery and agents before enabling cluster operations.
Use an existing Kubernetes platform to schedule the ClusterPilot API and workers with external PostgreSQL, durable shared artifacts, Secret references, optional Ingress, and HA safety controls.
Use the signed release bundle and Linux entrypoint for a controlled single-node profile with preflight-only, install, and verification modes.
Clear dependencies
You keep control of availability, encryption, backup, retention and recovery for every dependency.
Control-plane state, users, configuration, runs, audit, and projections
Run logs, command evidence, reports, bundles, and integrity metadata
32-byte protection key plus referenced database, registry, provider, and agent material
Digest-pinned API, agent, tools, Kubernetes, and add-on artifacts
NTP/local source, DNS, TLS, private CA or approved mTLS posture
Supported topologies
A profile defines placement. Backup and recovery tests are still required.
Evaluation or a service whose availability objective accepts one control-plane host. Durable state and tested backup remain mandatory.
Place API, workers, database, artifacts, and agents according to customer network and operational boundaries.
Requires external HA PostgreSQL, shared filesystem or S3 artifacts, leader election, ingress, and failure/recovery testing.
Before approval
Checksums, signatures, digests, SBOM, provenance, compatibility.
Database, artifacts, secrets, trust, clock, capacity, network.
Review effective configuration and apply through the supported surface.
Health, readiness, clock, build, replicas, leader, agents.
Prove the complete data, artifacts, key, and release recovery set.
Use a disposable provider/cluster evaluation with explicit acceptance.
Talk to ClusterPilot
We will compare your topology, registry, certificates, network and availability needs with the supported installation path.
hello@clusterpilot.de