Problemsby IbisTrouble with ClickHouse rows
Databases6 posts from 2 people3d active+00 posts in the last 7 days, 0 the 7 days before (fading)
Posts per day
The posts behind it
6, newest firstORDER BY f DESC LIMIT n with minmax index drops top rows when granule max is nan### Describe what's wrong - `ORDER BY f DESC LIMIT n` over a float column with a `minmax` index silently skips the true top rows. - Trigger: an index granule whose stored max is `nan` because one of its data granules is all-`nan` while another holds the largest finite values. -...ClickHouse/ClickHouse · clickgapaiOpen ↗TTL DELETE WHERE ignores session_timezone: DateTime string literal evaluated in server timezone, wrong rows silently deleted while the identical mutation predicate is rewritten### Describe what's wrong **TL;DR:** With `session_timezone` set, a `TTL ... DELETE WHERE` condition that compares a `DateTime` column with a bare string literal silently deletes the wrong rows: the literal is persisted verbatim in the table metadata and evaluated by background...ClickHouse/ClickHouse · zlareb1Open ↗max_rows_to_read / max_bytes_to_read silently ignored for file(), url(), s3() and object-storage reads### Describe what's wrong TL;DR: `max_rows_to_read`, `max_bytes_to_read` (and the read-speed / read-timeout / `*_leaf` limits) are silently ignored when a query reads through `file()`, `url()`, `s3()` and every other file / URL / object-storage table function or storage. The...ClickHouse/ClickHouse · zlareb1Open ↗Multi-shard DISTINCT with LIMIT silently drops rows on type collapse### Describe what's wrong - On a multi-shard Distributed table, `SELECT DISTINCT ts ... LIMIT 5` can return fewer distinct values than exist. - Trigger: shard column is `DateTime64(3)`, declared type on the Distributed table is `DateTime`, 10 distinct seconds underneath. -...ClickHouse/ClickHouse · clickgapaiOpen ↗`minmax` index and Parquet stats prune NaN rows under a negated predicate### Describe what's wrong - A negated float predicate silently loses the `NaN` rows. - On a `MergeTree` table with `INDEX f TYPE minmax`, `SELECT count() FROM t WHERE f != 1.5` returns 0 with the index and 13 without it; `NOT (f Analysis details (evidence, affected locations,...ClickHouse/ClickHouse · clickgapaiOpen ↗minmax skip index wrongly prunes rows with NaN in Map columns### Describe what's wrong - A minmax index on a `Map(..., Float64)` column drops matching rows instead of returning them. - Triggered by `SELECT count() FROM t WHERE m IN (map('a', nan))` with a minmax index on the Map column. - Expected count is 13 with `use_skip_indexes = 1`,...ClickHouse/ClickHouse · clickgapaiOpen ↗
Companies and products named
About this problem
Evidence
6 posts from 2 people in 1 place, about 2 a week over 15 days. Mostly on ClickHouse/ClickHouse. Tools named alongside: GitHub.
Frustration Frustration 0 of 3· Seen on GitHub issues
How it was grouped
Posts about ClickHouse with no known issue, sharing the word "rows". First post Sep 22, 2026, latest Oct 6, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method
Other ClickHouse problems
ClickHouse upgrades break working setupsTrouble with ClickHouse jsonClickHouse costs outgrow the value it givesClickHouse users keep filing bugsTrouble with ClickHouse columnExperiments on ClickHouse are hard to reproduce and trackTrouble with ClickHouse joinTrouble with ClickHouse nameTrouble with ClickHouse readClickHouse error messages are crypticTrouble with ClickHouse constantTrouble with ClickHouse slower
History
rising to fading
Trouble with ClickHouse rows (7 posts)