Overview.

We’re in an unprecedented time and going through the largest digital transformation in history, where enterprises are moving from using software to becoming software-driven.

This is a trend being manifested across all industries: financial services, retail, manufacturing, trans- portation and logistics, media and entertainment, healthcare, and many more.

To become a software-driven business that operates in real time, enterprises are undertaking multiple modernization initiatives, including application modernization, artificial intelligence and machine learning, the cloud, and edge analytics.

But modern applications no longer live in isolation.

They are built on microservices and rely on other services to move data between applica- tions.

Nowadays, modern distributed applications comprise hundreds of thousands of remote services operating in multiple tiers, all of which are loosely coupled and need to exchange data with each other.

In such distributed architectures, reliable data and message distribution is key.

Message-oriented middleware or messaging middleware has been widely used to handle the needs of message distribution and interservice communication across distributed applications.

Message-oriented middleware systems like enterprise service buses (ESBs), message queues (MQs), and data integration tooling for extract, transform, and load (ETL) act as the critical layer among services that need to communi- cate.

REST APIs have also emerged recently to enable simple point-to-point communication.

This paper takes a look at six Confluent customer implementations, the challenges posed by legacy approaches to messaging middleware, and the strategies they used to overcome them.