Install Camunda with Helm
Install Camunda 8 Self-Managed on Kubernetes using Helm charts. Choose the path that matches your environment and requirements.
Choose your installation path
| Use case | Installation guide | Secondary storage | Complexity | Best for |
|---|---|---|---|---|
| Testing & evaluation | Quick install | Embedded H2 (single-broker only) | Low | Local development, POCs, learning |
| Production with RDBMS | Install with RDBMS | PostgreSQL, Oracle, MariaDB, MySQL, SQL Server | Medium | RDBMS-first organizations, no ES/OS license |
| High-scale production | Install for production | Elasticsearch/OpenSearch or RDBMS | High | High-throughput, multi-team, OIDC, HA, monitoring |
- Use quick install if you want to evaluate quickly without external dependencies.
- Use install with RDBMS if your organization standardizes on relational databases.
- Use production install if you need security, scalability, and operational features.
- If you don’t have required infrastructure, see deploy required dependencies.
Installation guides
-
Quick install:
Deploy a single-broker Orchestration Cluster with embedded H2 secondary storage. Best for testing and early development. -
Install with RDBMS:
Configure Camunda to use a relational database for secondary storage, including schema setup and JDBC driver configuration. -
Production install:
Deploy a production-ready environment with TLS, OIDC authentication, multi-namespace support, and monitoring. Supports both Elasticsearch/OpenSearch and RDBMS backends. -
Deploy required dependencies:
Deploy PostgreSQL (management components), Elasticsearch/OpenSearch (secondary storage), and Keycloak on Kubernetes using official operators when managed services are not available.