Huolala Tech
Huolala Tech
Oct 13, 2022 · Big Data

How Druid Uses Bitmap Indexes for Fast Queries and Precise Deduplication

This article explains how Apache Druid builds and queries bitmap indexes for efficient OLAP analysis, and describes a dictionary‑encoding plus bitmap solution—adapted from Kuaishou—to achieve exact deduplication even on high‑cardinality dimensions.

Bitmap IndexDictionary EncodingDruid
0 likes · 14 min read
How Druid Uses Bitmap Indexes for Fast Queries and Precise Deduplication
ByteDance Data Platform
ByteDance Data Platform
Jan 24, 2022 · Databases

Accelerating ClickHouse LowCardinality: Merge Optimizations & Auto Fallback

This article details how ByteDance’s ClickHouse UBA edition improves dictionary encoding for low‑cardinality columns by redesigning the Part‑merge process, introducing a single‑dictionary merge, and implementing an automatic fallback for high‑cardinality columns, resulting in significant storage savings and query‑performance gains across large‑scale applications.

ClickHouseDictionary EncodingLowCardinality
0 likes · 12 min read
Accelerating ClickHouse LowCardinality: Merge Optimizations & Auto Fallback