Reading from Cassandra — internals and performance

Day 1 /  / Track 3  /  RU / For practicing engineers

This talk is about the detailed analysis of how the process of reading from the Apache Cassandra database takes place in terms of performance, what are the stages of processing requests, how long they take, which threads are responsible for what and how caching is organized. You will learn about the current stable version 3.11.x and also about the information on the read performance changes made in version 4.0.

The targeted audience — developers and administrators using Apache Cassandra. The talk will allow you to better understand how Cassandra works internally, what determines its read performance and what bottlenecks there are for this type of request.

Download presentation


Dmitry Konstantinov

A system architect and Java developer from Netcracker company. During the last 10 years participated in many different projects including network management, monitoring, and billing. Works a lot with open source technologies like Apache Cassandra, Zookeeper, Kafka, Hazelcast, etc. Professional interests: distributed systems, high performance, and high availability.

Invited Experts

Vladimir Sitnikov

Vladimir has been working on performance and scalability of NetCracker platform (a software used by telecom operators for automatization of network management processes and network equipment) for ten years. His focus is Java and Oracle Database performance optimisations. Vladimir is an author of more than a dozen of performance improvements in the official PostgreSQL JDBC driver.