July 2022 Summaries

We are pleased to welcome NANO Corp., a young French startup founded in 2019 by former defense personnel, as they share their journey from discovering ClickHouse to implementing it in production. NANO Corp.'s main goal is to analyze complex and changing network traffic at line-rate high speed using asynchronous feeds. They initially faced challenges with traditional RDBMS, including data schemes that relied on update speed and required clustering for performance issues. The company's pre-requisites were harsh, requiring top-notch performances over cost considerations. After exploring various options, they discovered ClickHouse, which met their needs due to its fast multiple inserts, indexing, materialized views, MergeTree function, and great documentation. Data ingestion in ClickHouse is more efficient, with a simpler architecture and less resource taxation, making it suitable for NANO Corp.'s technology, which involves sending data constantly while running constant queries on hot and cold data. The company stayed with ClickHouse despite other alternatives due to its perfect fit, overcoming their initial frustration with other databases.

Darwinium has chosen ClickHouse as its database engine due to its ability to simplify data pipelines, handle high-throughput writes and analytics of complex data types. ClickHouse's support for mutable engines, map and array data structures, and JSON type support make it an ideal choice for Darwinium's use cases. The platform can process a few thousand writes per second, with multiple data pipeline writers writing to a single server, and has consistent low CPU core readings. Additionally, ClickHouse supports cloud-native deployment, cost-aware storage patterns, and distributed and highly available architecture, making it well-suited for fraud and security domain use cases. The platform's rich function catalogue, SQL interface, and ability to deploy and run on various hardware configurations also make it a compelling choice.

DeepL is a company that uses ClickHouse as its central data warehouse to support various use cases such as analytics, company metrics, and technical monitoring. The company started using ClickHouse in 2020 to build up its analytics capabilities in a privacy-friendly manner, initially setting up a single node setup that proved capable of handling the amount of data they were throwing at it. They later invested heavily in automation, creating a combined source of truth for all events and table schemas, which enabled them to create complex events and queries that understood how users interacted with their platform. With this foundation, DeepL expanded its cluster from a single node setup to a more robust configuration, ingesting half a billion raw rows per day. They also utilized ClickHouse for experimentation, using it as the statistical analysis engine for AB-testing, which allowed them to rapidly iterate on frontend and algorithmic backend changes. Additionally, ClickHouse was used in their ML-infrastructure of personalization, serving as the data store for training and inference models based on user history.

At ClickHouse, we believe that Culture is above all being shaped and nurtured by our colleagues. It’s the way that we interact, work with each other, communicate and help each other learn along the way. Claire Lucas, Director of Global Business Strategy & Operations at ClickHouse, shared her experience on using language as a powerful tool to clarify intentions and set clear definitions in her interactions with team members across different languages and cultures. She also values thoughtful people working together, appreciates the well-engineered technology behind ClickHouse, and prioritizes making tools accessible to everyone, including wellness and nutrition coaching.

We're excited to welcome you to our upcoming events, including the ClickHouse v22.7 Release Webinar, in-person meetups in London, Munich, and Silicon Valley, as well as online events for those who can't attend in person. Our latest release, v22.6, includes several exciting features such as improved ARM support, dynamic values for search queries, and parameters for external user-defined functions. We've also published a Query of the Month article on geo queries for railway enthusiasts, which demonstrates how to find the closest stations to each other using ClickHouse's geo distance function. In our Reading Corner section, we've featured articles from Instana, Contentsquare, Cube, DENIC, QuickCheck, and others on topics such as optimizing complex queries, full-text search, ClickHouse scalability, and more.