We invite you to the next meeting of the Polish Java User Group in cooperation with Hazelcast (https://hazelcast.com/). The meetup will take place on 2nd of March from 18:00 in the Browar Lubicz (Lubicz 17J Street, room "Dojrzewalnia", level +1, https://browar-lubicz.com.pl/galerie/4,dojrzewalnia-wnetrza).
This time our speakers will be guests from Hazelcast - Jaromir Hamala (Director of Engineering) and Vladimir Schreiner (Product Manager).
The first presentation will concern the bare minimum for building reliable application, the second presentation will be about stream processing in Java.
For a detailed description of the presentations see below. Both presentations will be held in English.
During the meeting, we will have time for networking, and at the end of the meeting, we will draw prizes: software licenses and a book.
Hazelcast (https://hazelcast.com/) delivers the in-memory computing platform that empowers Global 2000 enterprises to deliver low-latency, data-centric applications. Built for ultra-fast processing at any scale, Hazelcast’s cloud-native, in-memory software accelerates business-critical applications.
Distributed Systems: The bare minimum for building reliable applications, Jaromir Hamala, Director of Engineering, Hazelcast
Distributed systems are hard. They introduce new failure modes, they force us to deal with consistency concerns. Researchers keep throwing scary acronyms such as CRDT, CAP or PACELC and we are forced to think about trade-offs all the time. What's considered to be a bug in one system can be a completely valid design in another system. It's easy to feel lost in all the intricacies.
This talk gives you a gentle introduction into consistency models and will help you with reasoning about trade-offs.
Stream Processing in Java, Vladimir Schreiner, Product Manager, Hazelcast
You want to collect metrics across your infrastructure, aggregate it and trigger an alert if the performance drops. In milliseconds. Which tools would you consider? The ELK stack? A timeseries database? A log aggregator?
There is no wrong answer. Fundamentally, what you really need is a query engine tuned to work with an append-only table - a stream processor.
Stream processors are building blocks for applications that run computations driven by the streams of data. We're going to discuss three focal points of modern streaming stacks: time, connectivity and scale.
The code samples used during the talk are build using Hazelcast Jet framework.