CloudNativePG provides production-grade HA PostgreSQL on Kubernetes but has no database branching, no developer self-service, and no org-wide RBAC. See how Vela fills those gaps as a BYOC alternative.
Last updated: March 2026
CloudNativePG (CNPG) is one of the best Kubernetes operators for running PostgreSQL in production — it handles primary/replica replication, automated failover, and object-storage backups well. But CNPG is an infrastructure primitive. It gives you a highly available database cluster; it does not give you database branching, developer self-service environments, or organization-wide RBAC. Teams that outgrow CNPG's primitives often look for a BYOC alternative that keeps their data in their own cloud while adding the developer-experience layer on top.
CloudNativePG is an open-source Kubernetes operator for running PostgreSQL clusters, created by EDB and a CNCF sandbox project. It uses a declarative CRD (Custom Resource Definition) approach: you define a Cluster resource and CNPG manages primary election, streaming replication to replicas, automated failover, and WAL archiving to S3/GCS/Azure Blob Storage. CNPG is Apache 2.0 licensed and widely regarded as the most actively maintained Kubernetes PostgreSQL operator as of 2026.
Best for: Platform engineering teams comfortable with Kubernetes who want open-source HA PostgreSQL and are willing to build developer workflow tooling on top.
How CloudNativePG (CNPG) compares to Vela BYOC across key dimensions
| Feature | CNPG | Vela BYOC |
|---|---|---|
| Deployment model | Kubernetes operator (CRDs) | BYOC — managed control plane in your cloud |
| High availability | Primary + replicas (streaming replication) | Built-in HA with live migration support |
| Automated failover | Yes — primary election on node failure | Yes — managed by control plane |
| Instant database cloning | No — full cluster required per environment | Yes — copy-on-write, any database size, seconds |
| Git-style DB branching | No — must implement yourself | Yes — branch per PR / pipeline / developer |
| Developer self-service | No — infra team provisions clusters | Yes — developers spin up DB branches via UI/API |
| Org-wide RBAC | Kubernetes RBAC (namespace-scoped) | Organization-wide RBAC across all databases |
| SSO / SAML / LDAP | Not included | Built-in SSO/SAML/LDAP integration |
| Backup & PITR | WAL archiving to S3/GCS/Azure (Barman) | Automated backups with configurable retention |
| Monitoring | Prometheus metrics exposed — you wire up dashboards | Built-in observability dashboard |
| Kubernetes expertise needed | High — CRDs, networking, storage classes | Low — abstracted by control plane |
| License | Apache 2.0 (open source) | Commercial (BYOC — data stays in your cloud) |
CloudNativePG (CNPG) is a CNCF sandbox open-source Kubernetes operator for PostgreSQL created by EDB. It handles HA clustering, streaming replication, automated failover, and WAL-based backups. Teams look for alternatives when they need capabilities CNPG doesn't provide: instant database cloning for dev/test environments, Git-style branching for CI/CD pipelines, developer self-service workflows, or organization-wide RBAC that goes beyond Kubernetes namespace-level access control.
CloudNativePG does not support database branching or instant cloning. To create an isolated copy of a database (for staging, testing, or a feature branch), you must provision an entirely new PostgreSQL cluster and restore from backup or run pg_dump/pg_restore. For a 100 GB database, this can take 30–60 minutes and incurs full storage costs for each copy. Copy-on-write cloning — as in Vela — provisions an instant branch of any database regardless of size, sharing unchanged storage blocks.
CloudNativePG requires significant Kubernetes expertise: you need to understand CRDs, storage classes, PersistentVolumeClaims, networking (Services, DNS), RBAC policies, and pod disruption budgets to run it well in production. Most teams need a dedicated platform engineer or SRE managing the operator. Vela's BYOC model deploys a control plane into your Kubernetes cluster that abstracts away this complexity — developers get self-service database environments without needing K8s knowledge.
If you need developer workflows (branching, instant cloning, self-service) on top of PostgreSQL in your own infrastructure: (1) Vela BYOC — adds copy-on-write database cloning, Git-style branching, and org-wide RBAC as a managed control plane in your cloud; (2) Neon OSS — self-hostable serverless Postgres with copy-on-write branching, but complex to operate; (3) Supabase OSS — full-stack backend with Postgres, requires managing 7+ services. If you just need production HA Postgres without the developer layer, CNPG, Zalando, or Crunchy PGO are all viable.
CNPG is suitable for enterprise teams with strong platform engineering, but it has gaps for enterprise governance: no org-wide RBAC beyond Kubernetes namespace isolation, no SSO/SAML/LDAP integration, no built-in audit logging for compliance, and no developer-facing self-service. Enterprises that need those capabilities typically layer tooling on top of CNPG — or choose a BYOC platform like Vela that includes them.
With CNPG, creating a test database for a CI/CD pipeline means spinning up a new cluster (5–15 minutes), restoring a dataset, running migrations, and tearing it down after. With Vela, a pipeline creates a copy-on-write branch of any production database in seconds via API, runs tests against it, and deletes the branch when done. Unchanged data blocks are shared — a branch of a 200 GB database uses near-zero additional storage until writes happen.
Keep your data in your own cloud. Add instant database cloning, Git-style branching, and org-wide RBAC — without replacing your infrastructure.