Case Study: Canva achieves 10x faster search and 70% lower costs with ClickHouse

How Canva achieved 10x faster search and 70% lower costs with ClickHouse

Canva needed a way to handle observability at massive scale as its user base grew, with nearly 3 million spans and 3 million logs generated every second. Using ClickHouse for tracing and logging, Canva set out to improve ingestion, storage efficiency, and search performance while keeping infrastructure simple and scalable.

ClickHouse helped Canva redesign ingestion with larger upstream batching, optimize schema and partitioning for trace searches, and add materialized views and query-layer improvements for flexible log search. The result was a 14x compression ratio, around 70% lower infrastructure costs, and a drop in P90 trace search time from 30 seconds to 2.5 seconds — a 10x performance improvement.

Canva

Zjan Carlo Turla

Software Engineer