Actual-time knowledge streaming and occasion processing are crucial elements of recent distributed methods architectures. Apache Kafka has emerged as a number one platform for constructing real-time knowledge pipelines and enabling asynchronous communication between microservices and purposes. Nevertheless, operating and managing Kafka clusters at scale will be difficult, requiring specialised experience and vital operational overhead.
Amazon Managed Streaming for Apache Kafka (Amazon MSK) is a totally managed service that lets you construct and run manufacturing Kafka purposes. With Amazon MSK, you’ll be able to depend on AWS to deal with the heavy lifting of provisioning and managing Kafka clusters, whilst you deal with constructing modern purposes and real-time knowledge processing pipelines.
On this publish, we discover how Fitch Group, one of many high credit standing corporations, used Amazon MSK and Amazon MSK Replicator to realize multi-Area resiliency for his or her mission-critical Kafka infrastructure.
About Fitch Group and their want for multi-region resiliency
As a number one international monetary info companies supplier, Fitch Group delivers important credit score and danger insights, strong knowledge, and dynamic instruments to champion extra environment friendly, clear monetary markets. With workers in over 30 nations, Fitch Group’s tradition of credibility, independence, and transparency is embedded all through its construction, which incorporates Fitch Scores, one of many world’s high three credit score rankings companies, and Fitch Options, a number one supplier of insights, knowledge, and analytics.
To remain aggressive and environment friendly within the fast-paced monetary business, Fitch Group strategically adopted an event-driven microservices structure. On the coronary heart of this ecosystem lies Kafka, particularly Amazon MSK, which serves because the spine for his or her knowledge integration methods.
Fitch Group makes use of Kafka to allow purposes to ship ratings-related enterprise occasions, facilitating automation inside their rankings workflow methods and offering real-time or close to real-time processing. This architectural selection has considerably lowered the time to marketplace for end-user-facing methods like Fitch Scores Professional and Fitch Group Scores web sites. Furthermore, Kafka’s strong capabilities enable for seamless aggregation and distribution of knowledge from many disparate methods by means of their knowledge platform, enhancing knowledge consistency, reliability, and accessibility throughout the group.
Given the crucial position that Kafka performs in Fitch Group structure, offering strong catastrophe restoration (DR) mechanisms turned paramount. Any disruption to their Kafka infrastructure may have vital repercussions on their rankings workflow automation, real-time processing, and end-user-facing methods, doubtlessly exposing Fitch Group to regulatory, monetary, and reputational dangers.
To attain the specified ranges of resiliency, Fitch Group had the next key necessities:
- Multi-Area deployment – Deploy MSK clusters throughout a number of AWS Areas to supply enterprise continuity and keep service availability throughout Regional or service occasions
- Automated replication – Replicate Kafka knowledge throughout Areas in close to actual time with minimal latency and knowledge loss
- Constant subject namespaces – Keep the identical Kafka subject names and constructions throughout supply and vacation spot clusters to reduce software adjustments
- Fast restoration – Within the occasion of a failover, allow purposes to seamlessly begin consuming from the replicated cluster with minimal Restoration Time Goal (RTO) and Restoration Level Goal (RPO)
Answer overview
Fitch Group selected to implement their multi-Area Kafka deployment utilizing Amazon MSK and MSK Replicator. MSK Replicator is a totally managed replication service that allows steady, automated knowledge replication between MSK clusters inside the identical Area or throughout completely different Areas. It helps replicating knowledge between clusters with completely different configurations, together with various dealer counts, storage volumes, and Kafka variations. Right here’s how Fitch Group used MSK Replicator to realize their multi-Area resiliency targets:
- Deployed MSK clusters in two separate Areas, with the first cluster in the primary Area and the secondary cluster in a special Area for catastrophe restoration
- Configured MSK Replicator to constantly replicate knowledge from the first cluster to the secondary cluster, sustaining the identical subject names and constructions throughout each clusters
- Carried out software failover logic to robotically swap to consuming from the secondary cluster in case of a major cluster unavailability, with minimal restoration time and knowledge loss
The next diagram illustrates this structure
Advantages achieved
By implementing Amazon MSK and MSK Replicator, Fitch Group realized a number of key advantages:
- Enhanced catastrophe restoration – The multi-Area deployment gives enterprise continuity even within the face of Regional or service occasions.
- Simplified operations – The managed functionality of MSK Replicator offloads the operational complexity of self-managing customized replication options, decreasing the burden on Fitch Group’s IT workforce
- Scalability – The answer can scale to deal with various knowledge masses, ensuring that DR capabilities develop alongside enterprise wants
- Minimal software adjustments – MSK Replicator helps replicating matters with the identical title, which eliminates the necessity for client software modifications, decreasing growth effort and potential errors
- Seamless failover and failback – Bidirectional replication capabilities allow fast switching of operations to the standby Area with minimal disruption, and simple reversion after the first Area is restored
- Improved testing capabilities – The setup facilitates common DR workouts with out impacting manufacturing methods, permitting Fitch Group to validate their DR plans constantly
Conclusion
By utilizing Amazon MSK and MSK Replicator, Fitch Group has efficiently carried out a extremely resilient and scalable Kafka infrastructure that meets their stringent enterprise continuity and catastrophe restoration necessities. This multi-Area deployment permits them to course of mission-critical monetary knowledge at scale whereas offering minimal downtime and knowledge loss within the occasion of service occasions or disasters. As Fitch Group continues to innovate and develop, their strong Kafka infrastructure gives a strong basis for future enlargement and the event of recent data-driven companies, in the end enhancing their means to ship well timed and correct monetary insights to their purchasers.
In regards to the authors
Kalyan Janaki is Senior Huge Information & Analytics Specialist with Amazon Net Providers. He helps prospects architect and construct extremely scalable, performant, and safe cloud-based options on AWS.
Venu Nemallikanti is the Enterprise Architect and Lead for Occasion Streaming at Fitch Group, a globally acknowledged monetary info companies supplier working in over 30 nations. His major obligations embrace overseeing the structure and implementation of occasion streaming options, making certain the seamless integration and efficiency of methods that ship credit score rankings, analysis, knowledge, and analytics to a worldwide clientele.
Chaitanya Shah is a Principal Technical Account Supervisor with AWS, based mostly out of New York. He likes to code and actively contributes to the AWS options labs to assist prospects remedy advanced issues. He gives steerage to AWS prospects on finest practices for his or her Cloud migrations. He’s additionally specialised in AWS knowledge switch and the information and analytics area.
Oleg Chugaev is a Principal Options Architect and Serverless evangelist with 20+ years in IT, holding a number of AWS certifications. At AWS, he drives prospects by means of their cloud transformation journeys by changing advanced challenges into actionable roadmaps for each technical and enterprise audiences.
