Openbank is leading a digital transformation in banking, leveraging public cloud technologies, microservices, and Continuous Delivery to deliver a superior customer experience.
Openbank, the digital bank of Santander Group, has one of the most complete, flexible, and agile digital banking platforms in the sector. It’s one of the first banks in the world to use a cloud-based IT infrastructure and offer a complete range of products (savings, loans, investments…) with customized, year-round customer service, 24×7.
Openbank is a modern cloud-native microservices and serverless application that runs on Amazon Web Services (AWS). The Openbank team found that scaling their existing CI/CD platform across tens of microservices, dev teams, and cloud environments became too complex and painful to manage.
CI/CD Before Harness
Like many organizations, Openbank historically built their own Continuous Delivery (CD) pipelines by extending Jenkins, their Continuous Integration (CI) platform. As their microservices and lambda architecture grew, managing those complex pipeline dependencies became brittle and painful to manage.
Specifically, a lack of insight into the deployment dependencies, status, and health during and post-deployment became very challenging. “Our Jenkins pipeline jobs were too painful to manage and deploy with,” said Javier Ros, Head of Architecture and Openbank. “We don’t trust Jenkins will always do the right thing.”
Deployments traditionally were centralized and managed by Operations, taking 3 to 15 engineers anywhere from 4 to 6 hours. Deployment verification and health checks could take as much as 2-3 hours depending on the service. Rollbacks were also manual and could take several hours to retrieve and sync the right artifacts, configuration, and environment setup.
Empowering Developers With Self-Service CD
Openbank’s CI/CD vision was to democratize deployments and empower its development teams with a self-service Continuous Delivery platform.
Javier and his team were tasked with looking at innovative solutions to enable this “shift left” from Operations to development. It was at this time Javier came across Harness.
Javier’s requirements for self-service Continuous Delivery were as follows:
- “Push button” and event-driven (e.g. webhook) deployment pipelines
- Provide deep integration with their AWS cloud stack (EC2, ECS, Lambda, CodeDeploy, KMS & CloudFormation)
- Provide real-time insight (single pane-of-glass) across all application, microservices, and deployments
- Reduce production incidents and increase code quality through canary deployments
- Auto-verify deployments using their existing monitoring stack of New Relic, Splunk, and CloudWatch
- Provide deployment and cloud abstraction so Cloud migration can be achieved in hours
The Openbank team started an evaluation of Harness less than 24 hours after the initial demo, and were up and running in days.
Reducing Deployment time and effort by up to 95%
Within a few weeks of implementing Harness, the Openbank team had reduced their average deployment time by 75% from 4-6 hours to 1-hour using a fully automated canary deployment workflow in Harness.
Instead of 3 to 15 operations engineers babysitting deployments, developers are now empowered to deploy their own code in less than 1 hour. That’s a productivity increase of 66% to 95% depending on the service and team. Development teams now achieve daily deployments on their own with true self-service capabilities.
Deployment verification and rollback are now fully automated using Harness’ Continuous Verification, which integrates directly with New Relic, Splunk, and AWS CloudWatch. Rollback time has dropped 96% from 2-3 hours to 5-10 minutes depending on the service.
The net result for Openbank is that they can deliver a superior customer experience. Not only does Openbank ship innovation daily to their customers, but they achieve this with minimal downtime or impact should any deployment fail.
Want to give Harness a test drive? Sign up for your free trial.