3 articles found
Architectural deep dive into ingesting real-time traffic data using Flink/Kafka pipelines and persisting low-latency states in Redis for sub-millisecond routing decisions.
A data engineer’s real-world postmortem on storing terabytes of hourly image data in Parquet files, and why the lakehouse dream becomes a performance nightmare
Apache Fluss 0.8.0 introduces Iceberg support, enabling sub-second latency processing via Flink and positioning itself as the hot layer solution for Iceberg-based lakehouses.
