Trayan Iliev

IPT – Intellectual Products & Technologies
iproduct https://www.facebook.com/trayan.iliev

Trayan Iliev (https://www.linkedin.com/in/trayaniliev/) is developer of end-to-end reactive full-stack applications and services using ES, TypeScript, Angular, React and Vue.js clients, and Spring 5, Reactor, Kafka, Pulsar, RabbitMQ, Express and Java EE back-ends. He has 15+ years of software development experience and 12+ years experience as enterprise IT trainer. Trayan is CTO of IPT – Intellectual Products & Technologies – training and consultancy company, focused on novelties in front-end and REST/gRPC/GraphQL backend development – reactive UI, end-to-end reactive programming, distributed event stream processing, real-time micro-service architectures, etc. Trayan has presentations on local developer conferences such as Voxxed Days, jPrime, jProfessionals, BGOUG, on topics like Spring 5, Reactor, SOA & REST, CDI, Java EE, React, Angular, Ionic, Node.js, RxJS, reactive java robotics, high-performance java. He is robotics / smart-things/ IoT enthusiast and organizes RoboLearn hackathons in Sofia.

Distributed Pub-Sub Messaging and Logging beyond Kafka with Apache Pulsar

Day 2 - 11th Dec 16:20-17:10 Hall 7 #CM Advanced Advanced

Apache Pulsar is an enterprise grade distributed pub-sub messaging platform, with a very flexible messaging model and an intuitive client API. It was initially developed by Yahoo and currently powers major Yahoo applications like Yahoo Mail, Finance, and Sports. It combines capabilities of a horizontally scalable distributed pub-pub messaging system (like RabbitMQ, ActiveMQ, ZeroMq, etc.), with that of persistent logging platform like Apache Kafka. The presentation introduces to Apache Pulsar and its major features:

  • horizontal scalability (millions of topics and millions of messages per second)
  • strong ordering and consistency guarantees
  • low latency durable storage
  • load balancing
  • multi-tenancy
  • authentication and authorization
  • quotas
  • supporting mix of very different workloads
  • optional hardware isolation
  • REST API for provisioning, admin and stats
  • geo replication
  • transparent partitioning
  • transparent message batching

Presentation finishes with a comparison between Kafka and Pulsar and a short demo.