Runink is a Go-native distributed pipeline orchestration and governance platform.

• Go + Linux primitives (cgroups, namespaces, pipes)
• Slice-based execution (ephemeral, secure workers)
• Raft-backed metadata stores for state consistency
• Schema contracts for trusted data evolution
• Built-in observability, security, and lineage

Runink empowers you to build declarative, auditable, and efficient data pipelines — without the complexity of Kubernetes or JVM-based stacks.

Curious? Detailed Architecture → | Components Overview → | Comparison with other open-source projects →

Key Concepts

	The golang code base to deploy features from configurations files deployed by command actions over the CLI/API.
Runink

	A single instance of a pipeline step running as an isolated Runi Slice Process managed by a Runi Agent within the constraints of a specific Herd
Runi

	A logical grouping construct, similar to a Kubernetes Namespace, enforced via RBAC policies and resource quotas. Provides multi-tenancy and domain isolation.
Herd

	A distributed, Raft-backed state store that guarantees strong consistency, high availability, and deterministic orchestration. No split-brain, no guesswork — just fault-tolerant operations.
Barn