PoME: Distributed Execution Proofs for Modern Microservices

Abstract

Microservices increasingly adopt cloud-native and distributed designs in a wide range of areas. However, the distributed nature of microservice architectures complicates trust and security verification across independent services, particularly in security-sensitive domains such as financial, healthcare, and government systems. This paper introduces PoME, a framework for Proof-of-Microservice-Execution (PoME) that enables verifiable and privacy-preserving computation in zero-trust environments. PoME integrates zero-knowledge proofs, blockchain, and a sidecar-based pattern to generate cryptographic attestations of microservice operations without revealing sensitive data. It supports end-to-end verifiable workflows through recursive proof aggregation and decentralized storage using the InterPlanetary File System for tamper-resistant records. Implemented and evaluated on a Kubernetes-based microservices benchmark, PoME demonstrates feasible performance with negligible latency overhead and scalable proof management. The results confirm that PoME effectively preserves bidirectional trust and auditability in distributed systems, offering a practical foundation for secure, verifiable microservice deployments.