Abstract
Building a scalable, multi-tenant backend for a Java-based, ML-driven Jira Cloud application imposes many requirements on the underlying technology stack. It is not uncommon to fulfill the requirements by combining pieces of technology—such as SQL and NoSQL databases, ORM tools, message brokers, load balancers, caching layers, ML pipelines, and web servers. In this talk, Peter shares his Apache Ignite experience. He will show how one can minimize the number of blocks in a complex, scalable backend for an ML-based, automated issue-management system (Alliedium), as you stay within the Java ecosystem and the microservice paradigm. We show you how to integrate Apache Ignite with Atlassian Connect Spring Boot, and we discuss Apache Ignite features such as SQL and NoSQL queries, thin and thick clients, caching, distributed messaging and events, distributed computations, database-schema change tracking, and Kubernetes deployment.