Handling Large Volumes of Immutable Structured Data With ClickHouse
summary
ClickHouse is shown handling 12 billion records daily on a small cluster and the demo uses a 48 million‑row sample imported in 2.5 minutes on an SSD (17 minutes on spinning disk), sustaining about 200,000 records/sec ingest. Raw data lives in MergeTree tables, aggregates are precomputed with materialized views into SummingMergeTree shards and a distributed table spreads inserts across nodes; aggregated queries return identical counts much faster and the setup works in production.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 0:37ClickHouse Background
- 2:52What is ClickHouse
- 3:48Online Analytical Processing
- 6:24Open Source
- 6:52C
- 8:15What I Like
- 11:40Table Definition
- 12:14Table Engine
- 14:24Summing Merge Tree
- 16:13Creating a materialized view
- 17:18Distributing data across a cluster
- 18:01rand
- 19:22command line
- 21:03aggregated data
- 21:48query count
- 22:13query on raw data
related talks
42mClickHouse: what is behind the fastest columnar database
34mClickHouse® - What is Behind the Fastest Columnar Database?
45mClickHouse: what is behind the fastest columnar database
17mBuilding Real-Time Data Lake with ClickHouse
51mWhat You Need to Know About ClickHouse Architecture
49mSolving real business cases with ClickHouse
30mBuilding For Fast
59mEffective Use of ClickHouse