Nava, a Brazilian infrastructure and monitoring company, migrated ELO's payments monitoring platform from Elasticsearch to ClickHouse, processing 22 million daily transactions. The migration reduced storage from 12 TB to 2 TB and cut annual infrastructure costs by 87% (from R$900,000 to R$120,000). Key technical optimizations included ZSTD/LZ4 compression, LowCardinality data types, materialized views, and projections. Query performance improved 5-9x depending on query type, and 300 real-time dashboards now refresh every five seconds with sub-2-second end-to-end latency. The shift from Elasticsearch Query Language to standard SQL also removed a developer bottleneck, enabling analysts to build dashboards independently.

8m read timeFrom clickhouse.com
Post cover image
Table of contents
How Lucas discovered ClickHouse #The limits of Elasticsearch #Migrating to ClickHouse #Cheaper, faster, and easier to use #Monitoring financial services across Brazil #

Sort: