Scaling Prometheus for Long-Term Storage with Thanos

Act as a Cloud Infrastructure Architect. Design a monolithic, highly available Thanos-based global query architecture specifically engineered to unify a massive, fragmented multi-region Prometheus setup anchored in [REGION]. Explicitly specify the technical configuration and deployment topology for the Thanos Sidecar, the Thanos Store Gateway, and the centralized Thanos Querier. Detail exactly how to securely bind the system to a highly durable object [STORAGE_BACKEND] (e.g., Amazon S3 or GCS) for infinite long-term retention. Architect and explain the mathematical downsampling process (Compactor) absolutely required to maintain sub-second Grafana query performance despite an overwhelming [QUERY_VOLUME] spanning years of historical data. Detail the strict gRPC mTLS communication settings required between components.