#Understanding distributed system optimizations
Martin Zapletal (zapletal_martin)
Cake Solutions
Performance is critical for many business use cases. This talk focuses on techniques for achieving optimized high performance in distributed systems while dependably providing required guarantees and ideally keeping the costs for resources, including CPUs, GPUs, memory and storage low.
Scalability and performance are critical for many business use cases. This talk focuses on techniques for achieving optimized high performance in distributed systems while dependably providing required guarantees and ideally keeping the costs for resources, including CPUs, GPUs, memory and storage low.
Firstly, the presentation introduces and in depth explains selected optimization techniques used in state of the art large scale stream and fast data processing frameworks such as Akka Streams, Spark or Flink, including configuration, deployment, logical and physical level optimizations. Consequently, powerful optimization concepts applicable to general distributed systems, including those built using Akka, will be presented and demonstrated on examples. Finally the role of machine learning and artificial intelligence in the optimizations area will be highlighted.
The attendees will gain understanding of the available optimization approaches and tradeoffs and ultimately will be able to apply some of the techniques to optimize their own system, whether general distributed systems written in Scala, systems based on Akka or any of the aforementioned technologies.
Martin is heading up Cake Solutions technical team in the US and is Apache Spark and Cassandra plugin for Akka Persistence contributor. Martin focuses on distributed systems, parallel and distributed approaches to data processing as well as machine learning, data mining in large volumes of data, and big data in general. These fields seem to be increasingly important in the industry and Martin has been promoting Scala, functional programming, and Reactive approaches as they provide very useful tools to solve these problems.