Industry: Technology
Tech stack simplification: from nine backend tools to one
Tencent integrated SurrealDB to unify its infrastructure monitoring stack. By leveraging SurrealDB's native graph capabilities, versioned data, and built-in analytics, Tencent replaced a fragmented system with a scalable, real-time monitoring platform.
Monitoring data lived across MySQL, Elasticsearch, Victoria Metrics, MongoDB, Doris, Trino, Rising Wave, Flink, and Dgraph.
Analysts struggled to pick the right storage or compute engines, increasing the platform team's maintenance burden.
Existing solutions lacked built-in graph views or recursive traversal for system relationships and time-based state transitions.
Scaling ingestion and queries for millions of time-series logs per day proved challenging.
Infrastructure components were modelled as graph nodes with edge-based relationships, enabling recursive traversal to trace event causality.
Snapshots and change-logs were stored using complex record IDs for efficient time-travel queries and temporal filtering.
Aggregated metrics were persisted into table views, reducing dashboard query latency and aiding policy enforcement.
Implemented LIVE SELECT on log tables for real-time alerts and diagnostics without polling.
SurrealDB replaced separate time-series, document, and graph stores, simplifying the stack.
Provided both historical snapshots and real-time streams from a single system.
Compound record keys and range filters handled millions of logs per day without schema changes.
Live queries powered alerting and UI refreshes, improving operational responsiveness.
SurrealDB's graph-native engine, versioned record access, complex IDs, record ranges, and LIVE SELECT queries provided Tencent with a unified and scalable foundation for infrastructure analytics and live telemetry.