Building a Highly Reliable Enterprise Infrastructure

In the digital payment landscape, the constant payment requirement from customers is driving the need for big financial enterprises like Citigroup to offer scalable and resilient infrastructure. Our mission-critical payment applications depend on Kafka’s distributed event streaming capabilities to connect new and legacy systems together, ensuring fast client payment processing speed and full compliance with regulatory requirements.

We must provide a high level of Kafka redundancy and scalability to support 99.9999% uptime. A highly available Kafka deployment must be able to survive a full single cluster outage while continuing to process events without data loss. Global payment applications not only require local high availability within one data center or one availability zone but across data centers as well. In addition, Kafka infrastructure must be able to handle very high payment throughput with scalability.

We will share how we:

-- drive the data streaming readiness by standardizing Kafka clusters among divergent payment application demands. -- overcome the challenge of designing and implementing Kafka enterprise infrastructure to meet business requirements -- stress test system stability under extreme conditions -- optimize for quick recovery from different failure scenarios