ClickHouse in the Data Lakehouse Architecture
Cloud ClickHouse is an open-source columnar relational DBMS for fast processing of analytical SQL queries on structured big data in real time. It is optionally included in the DLH architecture when real-time operational analytics need to be performed on the incoming event stream without the need to store data in S3 storage.
The component performs the following functions:
- Long-term data storage in a block-columnar format for performing BI analytics without queries to the main DLH storage.
- Processing complex read SQL queries for BI analytics with accelerated full scanning.
- Support for OLTP and OLAP scenarios within a single cluster DBMS.
- Support for data sharding and horizontal scaling up to 100 servers.
- Ability to directly import data and organize ETL and ELT processes from relational DBMS data sources.
- Continuous monitoring for performing operational analytics on the stream of incoming events from sources.
- The presence of ClickHouse in the architecture increases the cost of DLH, but at the same time allows launching a number of business scenarios in real time without deploying the full set of DLH components, which greatly reduces the resource intensity and T2M of the solution for large Enterprise installations.
Was this article helpful?