Everything you ever wanted to know about @ClickHouseDB but were afraid to ask... I interviewed the creator and CTO of ClickHouse, Alexey Milovidov and we discuss: • Why AI companies like OpenAI are adopting ClickHouse • How its architecture has evolved since 2009 • How

@ZachNai Maybe we do need Kafka after all!

@ZachNai And some cool visualisations as well. And one week to build it all. Not too bad!

The trades are coming into #ClickHouse Cloud

Little bit of Node, not too tricky! And now it's time for a demo

@ZachNai Spoiler!

@ZachNai From a web socket to ClickHouse without Kafka. Can we do it?

@ZachNai It's time to explore some market data!

@ZachNai Fast, resource efficient, and easy to use. Sounds good to me!

And now time for our last talk of the night by @zachnai

And it's 2/3 cheaper than it was with the old architecture. Winning all around

They had to balance the partition size to make sure the parts weren't too big or too small. But now it all works!

Now Josh is explaining the ReplacingMergeTree and techniques for deduplicating data - argMax and the FINAL keyword being competing approaches

@processout The business data model is in https://t.co/N2KDpWeBHt and we query ClickHouse further downstream

@processout ProtoBuf events from the event stream make their way into ClickHouse.

@processout And here's the new architecture! A much nicer diagram and it works better too.

@processout What led them to ClickHouse? Performance issues with Elastic!

And now it's time for Josh Thomas, who's gonna show us how @processout use ClickHouse

Cool things that ClickHouse can do.

ClickHouse London meetup getting started!