NetEase EasyData Metric Middle Platform: Architecture, Core Technologies, and Future Plans

NetEase EasyData (数帆) has evolved since 2006, using distributed databases, Hadoop, and launching products such as the big‑data platform Mengtian, NetEase YouShu, ChatBI, and the metric middle platform.

The product matrix includes low‑level data compute/storage components (HDFS/S3, Amoro, Yarn/K8s, Spark, Hive, Impala, Flink), DataOps lifecycle tools, and a nine‑module data governance suite covering standards, metadata, data maps, metric system, quality, asset center, model design, security, and services, plus BI/ML application layers.

The metric middle platform (EasyMetrics) addresses six common pain points: inconsistent metric definitions, fragmented entry points, unquantified metric value, low development efficiency, redundant calculations, and poor metric quality.

Solution features a unified logical semantic model layer that abstracts heterogeneous data sources, a custom metric query language that simplifies metric definition, composition, and time‑period handling, and a MetricsDSL that translates to Calcite RelNode and then to target‑engine SQL.

Core technologies include cross‑source logical semantic modeling, a concise metric query language (e.g., Select … where … BY …), time‑period syntax, and support for aggregation functions like AVG, COUNT, SUM, logical operators AND, OR, and arithmetic functions (+, -, *, %, ABS).

Engine decoupling enables flexible integration with third‑party schedulers (e.g., Apache DolphinScheduler) and compute engines (Spark, Flink, JDBC) via abstract adapters.

Future plans cover deeper metric applications such as dashboards, KPI management, and metric maps, broader BI integration, support for additional data sources like Doris, and AIGC‑driven natural‑language metric queries.