Healthcare Outbounds: Unbounded Scalability

HL7(Health Level Seven International)  enables healthcare system to exchange electronic health records (EHRs). Clinical and registration systems send HL7 data when a patient's demographics data or any other information is changed - such as  a patient is admitted, discharged or transferred(ADT) . We call this process as "outbound" transactions. It must adhere to the following rules -such as no duplicate transactions to be sent, order/sequence of transactions need to be maintained ( Suppose, patient A is admitted to a bed 101, he will be discharged and patient B will be admitted to the same bed, now the check-in transaction of 'B' should not go out before the Check-out transaction of patient 'A') and the processing time should be fast (what is fast? How do you measure it?)

Our multi JVM enterprise application had performance problem, as it used to process around 30,000 outbound transactions /hr. However, our big clients expect to process more than 65,000 transaction/hour, thereby it was necessary to re-design the system and meet the client goal. We built a micro-service based application to scale and process more than 90,000 transaction/hour. Apache Ignite and JMS  played a significant role in attaining the 90,000 scalability goal. In this talk Bruce and I will explain the problem with our database oriented architecture, the mistakes we made while redesigning the system with Apache Ignite, recommendations and learning

Technologies that will be covered: Apache Ignite's - In-memory data grid, Ignite SQL queries, Complex Event Processing and sending alerts, Java Messaging, Spring Framework.  The following  illities  will be covered - Resiliency,  availability  and High-performance

Purpose of the talk: To explain the importance of In-memory technology in healthcare system design. We will share the recipe used in healthcare domain to  redesign an existing (legacy) enterprise application for better scalability, resiliency and performance using Apache Ignite.

Target audience: Developers, architects and leaders who want know  - how Apache Ignite can scale your system and remove bottlenecks

What the audience will take away: Design concepts, refactoring legacy architecture using in-memory technologies and steps to come out from a monolith to a Microservice design