Amdocs Accelerates Cloud Growth for Fintech with DevOps-Led Migration
A DevOps-led migration to AWS, orchestrated by Amdocs, delivered the agility, availability, and security needed for fintech innovation and growth.
The fintech company needed to maximize resilience, availability and security while keeping costs in check. Amdocs achieved all this and more by orchestrating a containerized migration to AWS while implementing DevOps ways of working.
About the Customer: It is a scaling fintech offering CRM systems for mortgage, lending and banking providers. It specialises in the development of flexible, adaptable software that is easily tailored to support financial institutions’ existing business processes.
Migrating a monolith
Like most businesses entering the scale-up growth phase, the fintech needed to take steps to improve the resilience, availability and cost profile of its IT infrastructure. It was holding 211TB of data with a traditional managed hosting provider which resulted in a large monthly outlay. Day to day operations were manual, laborious and time consuming. What’s more, poor visibility and lack of access to the infrastructure meant that when outages occurred, they took a long time to resolve.
The fintech needed a better set-up, more suited to its growth ambitions. It had already containerized the infrastructure, and was looking for a partner to design, supply, build, test and commission a new cloud-based platform.
Amdocs was appointed to handle the entire migration process, and we recommended using AWS to host the new platform. This enabled the fintech to benefit from Amazon Elastic Kubernetes Service (EKS), a mature managed service for the automation, deployment, scaling and management of containers.
“It was so reassuring to see that Amdocs wasn’t phased by the scope and scale of our planned migration. This would have been a novel and high-risk undertaking for us, but Amdocs had the experience and insight to make it efficient and straightforward.”
Laying foundations for better performance
Before undertaking the migration, we held an Inception Day followed by a Discovery and Design session to ensure we had a solid understanding of the fintech’s requirements. After a successful pilot, we moved into the full migration phase. A key goal was to improve and modernize processes during the migration, so that the benefits of moving to the cloud could be realized quickly.
Our solution involved an everything-as-code approach across two platforms, one for production and one for non-production. They are held in separate accounts under a single AWS organization, with the non-production platform supporting static environments as well as dynamic, short-lived environments created via CI/CD pipelines.
Using AWS EKS enabled us to provide a resilient and scalable solution suitable to both production and non-production environments. It’s backed by RDS Aurora for MariaDB/MySQL hosting and the Elastic File System for file hosting. The EKS cluster is supported via an EC2 autoscaling group to provide the application hosting capacity, and it’s accessed via an Application Load Balancer.
To maximize resilience and availability, we spread services over two AWS availability zones (AZs). These are accessed via three gateways (as opposed to 15 gateways with the incumbent hosting provider). This achieved cost savings as well as enhancing security since the set-up makes it easy to whitelist IP addresses.
Access to services hosted on the Kubernetes environment is provided by an Application Load Balancer (ALB) connected to Ingress Services on the Kubernetes cluster. The cluster nodes are hosted in an AWS autoscaling group which spans the private zone. This provides resiliency across AZs and cluster capacity can be managed more intelligently, with autoscaling based on capacity requirements determined by the Kubernetes scheduler.
We used the MySQL-compatible AWS Aurora for high-performance database hosting. A key benefit of Aurora is its separation of compute and storage for better management of capacity. And Datadog was selected to handle metric collection and central log aggregation. This monitoring as a service solution helps with the correlation of data during events, reducing mean time to recovery (MTTR).
Finally, Terraform was used to deploy the AWS resources, with state stored on a secure S3 bucket. For the deployment of services such as cluster autoscaling and Datadog we used the Kubernetes package manager Helm.
DevOps principles facilitate scaling
The layout and technologies that we implemented during the migration have set the fintech up for future success. With a more resilient and adaptive network topology, it can develop its infrastructure in a sustainable way while meeting of-the-moment business needs.
While the cost-savings attached to the migration to AWS are significant, the long-term benefits go much further than this. We’ve replaced manual, error-prone processes with automated and consistent approaches. This gives the fintech’s developers more time to spend on the development of new products and features that enhance the customer experience and boost competitive differentiation.
Taking a DevOps approach to cloud migration has given the fintech the ability to strike the all-important balance between platform stability and rapid software development. The business is now ready and able to execute its growth plans, safe in the knowledge that the IT infrastructure can handle this exciting new phase.
“Amdocs has given us the confidence and capabilities to accelerate the fintech’s business progression. Our IT team has more capacity than before, and our developers have a new level of energy and enthusiasm now that they aren’t held back by infrastructure issues.”
Long-term value
The fintech’s strategic vision was for its in-house IT team to dedicate more effort to the continual improvement and innovation of products. Dovetailing DevOps principles with the migration to AWS has made this a reality. Going forward, Amdocs will continue supporting this through the provision of ongoing management, operation and maintenance of the new platforms.
