The guest article from ClickHouse evangelist Yegor Andreenko about an interesting extension of aggregate functions aggregate states, that can be used to pre-aggregate data that usually has to be kept in raw, like uniques.
We continue to benchmark ClickHouse against other analytic DBMSs. We were inspired by the benchmark with star2002 experiment dataset described here, and decided to replicate it using ClickHouse. That gives another interesting comparison vs Amazon RedShift.
ClickHouse is an exellent DBMS with very smart people working on it. Unfortunately, it still lacks some important communication procedures, and arguable the most wanted one is release notes.
Altinity mission is to make ClickHouse use easy for everbody, and we will try to fill the gap between Yandex and end users.
Below are release notes that Yandex team presented at the recent meetup in Ekaterinburg. It includes changes and new features appeared in latest 1.1.54231 release as well as some others that were not yet documented or publically explained.
There are few ClickHouse benchmarks in the web already. Most of them use denormalized database schema. However, in denormalization is not always possible or desirable. In this article we will compare the query performance between denormalized and normalized schema where normalization is modelled using unique ClickHouse external dictionaries feature.