Hello! On Thursday, January 23, we will have our first meetup about Apache Kafka. We will discuss integration, source connectors, and launching scalable applications in Kubernetes. Presentations will be made by engineers from Booking.com, Confluent, and Avito.
We meet at 6:00 PM. Register yourself and invite colleagues. Below — announcements of the talks, a registration link, and a video stream of the meetup.
Reports
Connectors: why are there so many of you? — Anatoly Soldatov, Avito
Recently, Martin Kleppmann commented on Twitter about Netflix's article on their new CDC connector: "Netflix created their own Change Data Capture (CDC) connector for MySQL and PostgreSQL… although perhaps it would have been better to contribute to Debezium?" In my presentation, I will give a brief overview of source connectors. We will then look at the advantages and disadvantages of well-known connectors for PostgreSQL, including Debezium. Finally, I will try to answer Martin's tweet and explain why we decided to create our own CDC connector at Avito and what resulted from it.
How to tame Apache Kafka? The experience of Booking.com — Alexander Mironov, Booking.com
Despite all the capabilities of Apache Kafka, it is not easy to integrate and maintain. In the presentation, I will explain how we solve this challenge at Booking.com, where Kafka and the service built around it are used as the main product for streaming data. We will take a detailed look at how a relatively small team maintains control over dozens of clusters and ensures data security, while also sharing laughs — or tears — over the bugs and issues we have had to tackle.
Kafka on Kubernetes: is it Juan's sombrero? — Victor Gamov, Confluent
If you need to choose the most efficient distributed platform for processing real-time streaming data, the answer today is obvious — it's Apache Kafka! If you need to deploy a scalable application in the cloud with high resilience, redundancy, load distribution, and containers, you need Kubernetes.
And if you cross this snake and hedgehog, you get a scalable application processing real-time streams in the cloud. Great, isn't it? But is it all that simple? Are there enough Lego bricks that basic Kubernetes provides to launch serious distributed systems that require state storage?
I am Viktor Gamov, a Developer Advocate from Confluent, and I'll help you navigate through this complexity. I'll also attempt to answer the question, 'Is it worth running Apache Kafka and its components, as well as stream processing applications on Kubernetes?'
Passwords and credentials
The meetup will start on January 23 at 18:00. Participation is free, but you need to . Please provide your real first and last name in the registration form. Don't forget to bring your passport or driver's license to get a pass at the reception on the day of the event.
Address: Avito Office, .
If you cannot attend the meeting, you can watch the live stream on the . You will be able to ask questions during the live broadcast, which we will convey to the speakers.
See you there!
Source: habr.com
