--- title: "High Cardinality in ClickHouse at Scale: What Actually Breaks" url: https://daily.dev/posts/high-cardinality-in-clickhouse-at-scale-what-actually-breaks-pc6axtguq source_url: https://last9.io/blog/clickhouse-high-cardinality-at-scale type: article source: "Last9" published: 2026-06-25T06:36:58.325Z updated: 2026-06-25T06:37:24.828Z tags: ["observability", "clickhouse"] reading_time: 13 upvotes: 0 comments: 0 language: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # High Cardinality in ClickHouse at Scale: What Actually Breaks **[Last9](https://daily.dev/sources/last9)** · 13 min read · 0 upvotes · 0 comments ## Summary Running ClickHouse at scale with high-cardinality telemetry data reveals a deceptive failure pattern: ingest stays healthy while query-time costs accumulate silently. Three distinct failure modes emerge at scale — memory exhaustion during GROUP BY aggregations, part explosion from fine-grained grouping keys in materialized views, and full-text scans on unindexed high-cardinality columns. Real production numbers are shared, including a spike to 400M rows/minute that caused 160 concurrent merges and 40-50 second insert latencies. Mitigations include designing ORDER BY keys around real query patterns, adding tokenbf_v1 bloom filters for high-cardinality text lookups, coarsening timestamp granularity in materialized views, promoting hot Map keys to typed LowCardinality columns, and avoiding SELECT * on wide tables. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [High Cardinality Metrics: How Prometheus and ClickHouse Handle Scale](https://daily.dev/posts/high-cardinality-metrics-how-prometheus-and-clickhouse-handle-scale-o0heii5ky) · Last9 · 1 upvotes · 0 comments - [Why High-Cardinality Metrics Break Everything](https://daily.dev/posts/why-high-cardinality-metrics-break-everything-ohhcpvwbw) · Last9 · 12 upvotes · 1 comments - [The InfluxDB Cardinality Problem: Why High-Cardinality Industrial Data Breaks It](https://daily.dev/posts/the-influxdb-cardinality-problem-why-high-cardinality-industrial-data-breaks-it-szpfxjauu) · CrateDB · 0 upvotes · 0 comments --- Tags: [#observability](https://daily.dev/tags/observability), [#clickhouse](https://daily.dev/tags/clickhouse) [View this post on daily.dev](https://daily.dev/posts/high-cardinality-in-clickhouse-at-scale-what-actually-breaks-pc6axtguq)