Inclavare Containers is a CNCF sandbox project providing a confidential container runtime for cloud-native confidential computing. It enables running sensitive workloads in hardware-assisted Trusted Execution Environments (TEEs) like Intel SGX, isolating them from untrusted infrastructure providers and insiders. The project aims to simplify the adoption of confidential computing for cloud-native applications.

How It Works

Inclavare Containers leverages OCI-compliant components, including rune and shim-rune, to interface with backend enclave runtimes like Occlum (a Library OS for Intel SGX) or WAMR. This architecture allows for hardware-enforced isolation, removing the Cloud Service Provider (CSP) from the tenant's Trusted Computing Base (TCB). It also includes an Enclave Attestation Architecture (EAA) with Rats-TLS for secure, cross-platform remote attestation, proving workload integrity.

Quick Start & Requirements

• Installation: Build from source using make within a Docker development environment. Pre-built RPM/DEB packages are also available.
• Prerequisites: Requires hardware TEE support (e.g., Intel SGX), corresponding SDKs, and specific drivers. Building requires Docker.
• Integration: Can be integrated with containerd, dockerd, and pouchd.
• Resources: Building requires a development environment with SGX setup. Deployment details vary based on the chosen enclave runtime (Occlum, WAMR).
• Documentation: Official Quick Start, Architecture, Roadmap.

Highlighted Details

• OCI-compliant runtime (rune) and containerd shim v2 support (shim-rune).
• Universal, cross-platform remote attestation infrastructure (EAA) with Rats-TLS.
• Supports multiple enclave runtimes, including Occlum (Library OS for SGX) and WAMR.
• Designed for cloud platform agnosticism, deployable on any public cloud Kubernetes.

Maintenance & Community

• CNCF sandbox project.
• Development primarily driven by Alibaba.
• Roadmap available for future development plans.

Licensing & Compatibility

• The project's license is not explicitly stated in the README.

Limitations & Caveats

• Building and running requires specific hardware TEE capabilities (e.g., Intel SGX) and associated software stacks, which can be complex to set up.
• The README does not specify the project's license, which is crucial for commercial adoption.