ClickHouse

Your favorite Postgres database now comes with a quest log 🎮 We redesigned ClickHouse Managed Postgres onboarding around the fastest path from an empty service to doing something useful with your data. Your objectives, should you choose to accept them: 1️⃣ Spin up a Postgres service 2️⃣ Migrate or ingest data 3️⃣ Run your first query 4️⃣ Run analytics with ClickHouse 🔓 Bonus level: use pg_clickhouse as one SQL layer across both engines. https://lnkd.in/eEG3dNUJ

QuintoAndar is the largest housing platform in Latin America, closing 15,000 new rental contracts every month across 60,000 active listings. Their customer data platform (CDP) collects over 1 billion events a month from 14 million users. When the two-pipeline setup behind it became too slow and expensive, they turned to ClickHouse Cloud. Then they hit a wall. Marketing wanted to use Hightouch’s Customer Studio feature to build audiences and journeys, but it doesn’t support ClickHouse as a source. They needed a way to put a supported source in front of ClickHouse… and found their answer in managed Postgres. Tech lead manager Bruno Brito joined us on a recent webinar, where he walked through the rebuild, the roadblock, and why Postgres managed by ClickHouse turned out to be the missing piece of their architecture. Read the full story 👇 clickhou.se/4zBrdyd

Last chance to register! Next week (Tue, Sept 1), our Open House Roadshow in EMEA kicks off in Amsterdam, and seats are almost gone. You'll get: keynotes, hands-on training, and customer stories from Picsart, Trade Republic, and Infinitas Learning, all straight from the engineers building ClickHouse. We have 20 more seats left, make sure to secure your spot! 👉 Register now: https://lnkd.in/e58gHxUU #ClickHouse #OpenHouse #Amsterdam

High cardinality is where Prometheus starts to hurt, and it comes down to how it stores data: every unique combination of labels becomes its own time series, with its own memory, index, and lifecycle. That adds up fast. A thousand hosts times a hundred apps times five status codes times fifty endpoints is 25 million series, and each one carries around 200 bytes just for its in-memory struct, before any samples. Add one more label like container_id and the count multiplies again, which is how teams end up with Prometheus using tens or hundreds of gigabytes of RAM. https://lnkd.in/etaqmVrq

ClickHouse is very proud to have our own Data Scientist Ally (McNulty) Godine speaking in New York about our AI-first internal data warehouse that she uses on a daily basis. This will be a great session to learn about how you can leverage AI in your own workflows. To join us in person RSVP here: https://luma.com/xf3wghdh

Engineers across ClickHouse use Claude Code to ship faster. "We built our own AI agents in ClickHouse that teams interact with directly, including an agent in the SQL console and an AI SRE. We use Claude Code to build and iterate on these agents themselves. The tooling that powers our customers' AI experiences is, in part, built with AI" - Alexey Milovidov "The prototype for clickhousectl was built in a few hours using Claude Code. Four months later, we've shipped 26 versions of clickhousectl, where 99% of the code is written by Claude Code." - Alasdair Brown

We've seen a lot of conversation around the "observability wars" and the idea that ClickHouse might be winning them. We're humbled, but declaring victory would be premature. ClickHouse has become a common foundation for observability products, particularly for storing and querying telemetry at scale. We're proud of that. But a database is one layer of an observability system. Collection, workflows, visualization, alerting, and investigation matter too, and there's plenty we still need to build. For observability as a whole, the work isn't done. https://lnkd.in/e6YYrcX4

A 512 KiB read buffer was quietly sending every query response to .NET’s large object heap. 🧹 ClickHouse.Driver 1.3 lowers the default to 8 KiB, alongside a broader set of improvements delivered since 1.0. - Over 80% lower per-query allocation across 1,000 small SELECTs - Gen 2 collections eliminated in the same benchmark - More than 90% less total GC pause time - Type-safe POCO reads and writes - Official EF Core support, including bulk inserts and migrations https://lnkd.in/eqPsnt93

We’re partnering with Striim to give enterprises a continuous path from Oracle, SQL Server, Snowflake, and other operational systems into ClickHouse. Striim captures inserts, updates, and deletes as they happen, processes them in flight, and delivers them directly into ClickHouse for analytics, applications, observability, and AI. Together, we’re closing the gap between when something changes in the business and when that change becomes available for analysis.