Modern telemetry systems require both highthroughput time-series ingestion and flexible structured query
capabilities. This paper presents a comprehensive evaluation of
a dual-store telemetry backbone combining QuestDB for timeseries workloads and CrateDB for structured analytics. Through
systematic benchmarking of ingress throughput, query latency,
and retention costs, we demonstrate that QuestDB achieves
40% higher ingestion rates (45K vs 32K records/sec) with 33%
lower P95 latency (12.5ms vs 18.7ms), while CrateDB provides
superior structured query flexibility for complex analytics. Our
dual-store architecture achieves 38K records/sec with 15.8ms
P95 latency, representing an optimal balance for heterogeneous
telemetry workloads. Cost analysis reveals QuestDB’s 30% storage efficiency advantage for long-term retention, while CrateDB’s
structured indexing provides 2.5x faster complex query performance. These findings inform architecture decisions for largescale telemetry systems requiring both real-time streaming and
analytical capabilities.