Harbor is an open-source, cloud-native container registry that extends the open-source Docker Distribution with enterprise-grade features. It provides integrated vulnerability scanning, identity management via LDAP/OIDC, and policy-based image replication across geographically distributed instances, serving as a secure, compliant artifact repository for container images and OCI artifacts like Helm charts.
Glossary
Harbor

What is Harbor?
Harbor is an open-source, cloud-native container registry that extends the Docker Distribution with integrated security scanning, identity federation, and policy-driven replication.
Operating as a private registry, Harbor enforces content trust through integration with Notary and Cosign, cryptographically validating image signatures before deployment. Its built-in garbage collection and retention policies manage storage quotas, while geo-replication ensures high availability. Harbor's admission controller integrations enable Kubernetes-native enforcement, preventing unsigned or vulnerable images from being instantiated in production clusters.
Core Capabilities of Harbor
Harbor extends the open-source Docker Distribution with security scanning, identity management, and policy-based replication, providing a production-grade private registry for cloud-native AI workloads.
Identity Federation and RBAC
Harbor decouples authentication from the registry itself, supporting integration with LDAP, Active Directory, and OIDC providers. Once authenticated, fine-grained role-based access control governs every operation:
- Project-level isolation: Each project is a namespace with its own user, robot account, and role assignments
- Robot accounts: Machine identities with narrowly scoped permissions for CI/CD pipelines
- Audit logging: Every pull, push, and delete action is recorded for compliance
OCI-Compliant Artifact Storage
Beyond container images, Harbor functions as a generic OCI-compliant artifact repository. It can store and serve any content type that adheres to the OCI distribution specification, including:
- Helm charts: Versioned Kubernetes application packages
- Cosign signatures: Cryptographic attestations stored alongside images
- SBOMs and provenance: Supply chain metadata attached as referrer artifacts This unified storage model eliminates the need for separate artifact registries.
Garbage Collection and Retention
Harbor provides automated storage reclamation through garbage collection and tag retention policies. These mechanisms prevent unbounded storage growth in production registries:
- Garbage collection: Removes untagged manifests and unreferenced layer blobs
- Retention policies: Automatically delete images based on tag count, age, or vulnerability severity
- Quota management: Enforce per-project storage limits to prevent resource exhaustion
Proxy Cache for Air-Gapped Environments
Harbor can operate as a pull-through proxy cache, mirroring upstream registries like Docker Hub or Quay.io. This capability is essential for air-gapped or bandwidth-constrained deployments:
- Local caching: First pull fetches from upstream; subsequent pulls serve from local cache
- Offline operation: Pre-populate the cache before disconnecting from external networks
- Bandwidth reduction: Eliminate redundant external downloads across large node fleets
Frequently Asked Questions
Clear answers to common questions about Harbor, the open-source cloud-native registry for securely storing, scanning, and signing container images and OCI artifacts within sovereign infrastructure.
Harbor is an open-source, cloud-native container registry that extends the open-source Docker Distribution with enterprise-grade features including vulnerability scanning, identity management, and policy-based image replication. It functions as a private artifact repository that stores, signs, and scans container images and OCI-compliant artifacts. Harbor deploys as a set of Docker containers orchestrated by Docker Compose or Kubernetes, with a PostgreSQL database for metadata, Redis for job queuing, and a Trivy or Clair scanner for vulnerability analysis. It proxies the upstream Docker Distribution API while adding a robust authentication layer, a web portal for management, and a comprehensive REST API for automation. When an image is pushed, Harbor automatically triggers a vulnerability scan against known CVE databases, enforces image signing policies via Notary or Cosign integration, and can replicate the artifact to remote registry instances based on configured replication rules. This architecture ensures that only trusted, scanned, and compliant artifacts enter production environments.
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.

Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.

Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
Related Terms
Explore the critical components that integrate with Harbor to form a complete, secure, and sovereign container artifact management strategy.
SBOM
A Software Bill of Materials is a nested inventory of all components, libraries, and dependencies within a software artifact. Harbor can generate and store SBOMs in SPDX or CycloneDX formats, providing deep supply chain transparency. This allows security teams to quickly identify if a newly disclosed vulnerability (like a Log4Shell variant) impacts any image stored in the registry by querying the SBOM rather than rescanning every layer.
Admission Controller
A Kubernetes-native plugin that intercepts API requests and can validate or mutate objects before they are persisted. When integrated with Harbor, an admission controller can enforce policies such as:
- Only allow images from a specific Harbor project
- Block images with critical vulnerabilities
- Require valid Cosign signatures This prevents non-compliant workloads from ever being scheduled on the cluster.
Registry Mirror
A local, read-only replica of an upstream registry that serves as a pull-through cache. In air-gapped or bandwidth-constrained environments, a Harbor instance can be configured as a proxy cache for registries like Docker Hub or Quay.io. This drastically reduces external bandwidth consumption and ensures that images remain available for pulling even if the upstream source is temporarily unreachable.
SLSA Provenance
A security framework (Supply-chain Levels for Software Artifacts) that provides a tamper-proof, attestable record of an artifact's build process. Harbor stores SLSA attestations as OCI artifacts, linking each image to verifiable metadata about its origin—including the builder, source repository, and build steps. This creates an auditable chain of custody from source code to deployed container, critical for meeting compliance requirements.

About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us