How JD.com Built a Lakestream with Apache Fluss and Apache Hudi
JD.com keeps fresh data in Fluss and tiers it into Hudi, cutting batch compute cost by more than 80% and storage cost by more than 50% in recommendation ranking.
The Fluss Blog.
Engineering, ideas, and stories from the streaming frontier.
34 articles
Crash-safe retries for side-effecting workers with Apache Fluss
Recover a side-effecting worker from Fluss event history, reconcile uncertain actions with downstream receipts, and query the trace with Flink SQL.
Apache Doris x Fluss: Unified Queries Across Stream and Lake
How Apache Doris 5.0 reads Fluss log and primary-key tables, combines Paimon history with recent data, and manages JVM memory during Union Read.
The Streamhouse And Lakestream With Apache Fluss
Shared tables for streaming, serving and analytics
Announcing Apache Fluss 1.0: Real-Time Data Foundation for AI
Fluss 1.0 strengthens the real-time data foundation for AI with Fluss Gateway, native clients, and deeper streaming and lakehouse integration.
From Kafka to Fluss: How Rednote Migrated a Core Real-Time Indexing Pipeline
A production case study in columnar streaming, cold-data isolation, and lakehouse integration.
Apache Fluss Graduates to a Top Level Project
Apache Fluss graduates to a Top-Level Project, marking a milestone in community growth and the evolution of real-time Lakehouse infrastructure.
Tiering Service Deep Dive Part 3: In Production
Run lake tiering in production: identify failure modes, avoid deployment pitfalls, and monitor the signals that reveal operational health.
Tiering Service Deep Dive Part 2: Tuning
Tune lake tiering with a practical guide to parallelism, table types, freshness, multi-table scheduling, and scaling out tiering jobs.
Tiering Service Deep Dive Part 1: The Mental Model
Follow a lake-tiering round from its timer to the committed snapshot, and understand the processes and table states that make it work.
Showing 9 of 34