The OLAP Strategist: Druid Edition

Druid Architecture Notes

Navigating the trade-offs of real-time analytics, distributed ingestion, and high-concurrency query execution.

Read notes Apache Druid
Latest notes

Operational Druid topics, kept compact.

2021-08-15

Speed Up Query Execution Using Vectorization

Vectorized execution in Apache Druid processes row batches instead of single rows, reducing method-call overhead and improving CPU/cache efficiency.

GroupByTimeseriesVectorizeDruid 0.20.2
 2021-08-15

Auto Scaler Kafka Ingestion Tasks

A short reference note for dynamic auto-scaling of Kafka stream ingestion tasks in Apache Druid.

KafkaIngestionAutoscalingDruid
Study map

How the notes fit together.

Query execution

Vectorization, segment scanning, context parameters, and engine support boundaries.

Ingestion operations

Kafka task scaling, stream load changes, and supervisor-level operational behavior.

OLAP serving

Where Druid fits in low-latency analytics stacks and dashboard-facing serving layers.

Production tuning

Small, specific notes that turn release details into operational decisions.