CASE STUDY:

Mirantis helps Asian bank scale mobile banking for millions of customers during COVID-19 pandemic

The Business

An Asian bank with millions of customers is the top commercial lender in its country and serves the majority of the nation’s households. The bank operates hundreds of branches nationwide and was the first bank to introduce many fin-tech advances in its market, including mobile banking, internet banking, digital applications, and kiosks. Trusted by customers for its strong financial performance and reliability, the bank has won many industry awards for its services.

Company

  • Commercial and consumer bank

  • Millions of customers

  • Asia HQ

Surging Customer Demand

Like most lenders around the world, the bank had been investing in digital transformation to reduce manual processes that resulted in human errors and long wait times at branch offices. As the bank introduced digital banking services, it began using VMware to deploy monolithic banking applications on bare metal in its on-premises data centers. With the popularity of its mobile apps, the bank’s customer base quickly grew 400%, but traditional virtualization didn’t provide the scalability it needed to accommodate surging customer demand. Additionally, the lender wanted to reduce downtime for its services.

“The challenge with our traditional IT structure was not being able to handle those increased loads,” said a senior automation engineer at the bank, who leads the company’s DevOps team. “To give access to millions of customers, we needed a very scalable, highly available system, and one of the main ways to resolve this issue was to go cloud native,” he said.

“To give access to our millions of customers, we needed a very scalable, highly available system, and one of the main ways to resolve this issue was to go cloud native.”

Sr. Automation Engineer

Enterprise Kubernetes

The bank conducted a proof of concept (POC) with several top Kubernetes vendors, and found Mirantis Kubernetes Engine best suited its needs, especially Mirantis’ enterprise support options. Mirantis Kubernetes Engine (formerly Docker Enterprise) is an enterprise Kubernetes platform that provides the fastest and easiest way to deploy cloud native applications at scale in any environment. The bank integrated Mirantis Kubernetes Engine with Azure DevOps for source control, Jenkins for build automation, and JFrog for security analysis and vulnerability scanning. They soon began migrating legacy applications to containers and microservices.

As with most organizations, moving from virtualization to containerization involved a major cultural shift, but urgent business needs compelled senior management to accept change.

“Most directors have a more traditional way of thinking and were hesitating with such a big change,” the senior automation engineer said. “What convinced them was the rapid increase in customers that we were expecting, and the downtime we were experiencing. It was obvious that traditional infrastructure couldn’t scale up with the growing demand from customers.”

Digital Acceleration with COVID-19

A few months after the bank began using Mirantis Kubernetes Engine, the COVID-19 pandemic hit. Customers no longer wanted to visit branch offices in person, and demand for mobile and online services soared. As the government implemented lockdown restrictions, the bank’s team of developers was able to rapidly respond to changing needs due to their adoption of cloud native technologies. The bank quickly launched new services and functionality for mobile and Internet banking, call centers and kiosks around the country. In 2021, the bank revised its digital strategy, pushing to digitize all business processes and upgrade its network of thousands of ATMs, hundreds of express banks, and hundreds of kiosk machines.

One of the bank’s most challenging days of the pandemic occurred when the federal government issued a COVID-19 relief payment to millions of citizens in the country. On the day it was distributed, thousands of customers rushed to cash out, causing traffic to spike 300% more than normal, but because applications were containerized, the bank was able to swiftly scale out resources to accommodate peak demand.

“There were huge lines at the ATM machines and branches that day, and it really took a toll on our infrastructure, but we handled it very well,” the senior automation engineer said.

“With Kubernetes and containers, if there’s one specific service that needs a lot more resources than usual, that certain service can scale up, while others can be downgraded to use less resources, thus keeping the overall infrastructure stable,” he added.

“With Kubernetes and containers, if there’s one specific service that needs a lot more resources than usual, that certain service can scale up, while others can be downgraded to use less resources, thus keeping the overall infrastructure stable.”

— Sr. Automation Engineer

Delivering Services without Interruption

Already, the bank has containerized roughly 90% of its applications, and the whole IT department is using Kubernetes. Currently, the lender runs dozens of applications with hundreds of shared microservices in production on MIrantis Kubernetes Engine. These include business critical applications for money transfers, loan requests, e-commerce, credit/debit card transactions, and in-bank shopping. Using containers has also enabled the bank to adopt new use cases, such as machine learning for fraud detection and approval automation for small loans.

Whereas with virtual machines, the bank would typically deploy applications only once every 1-2 weeks, now they have the technical capability to deploy daily, without any downtime. Since adopting containers and Kubernetes, the bank has seen downtime drop by roughly 70%, despite supporting significantly more services and 400% the number of customers.

Resource utilization has also greatly improved. The bank currently runs hundreds of microservices in production on only 20 nodes. In the past with virtual machines, this would have required at least 50 physical machines, the senior automation engineer estimates.

All of these benefits have enabled the bank to deliver services without interruption to customers despite COVID-19 lockdown restrictions. In 2021, the bank’s digital platform adoption reached an all-time high with millions of users, and almost 100% of all transactions were completed through digital channels.

Challenges

  • Increase scalability to accommodate 400% customer growth

  • Improve production uptime of applications and services

  • Boost agility to respond rapidly to the pandemic emergency

Solution

  • Mirantis Kubernetes Engine with enterprise support

  • Azure DevOps version control

  • Jenkins build automation

  • JFrog security analysis and vulnerability scanning

Results

  • Greater scalability to serve sudden spike in users

  • Technical capability to deploy daily, instead of once every 1-2 weeks

  • Roughly 70% decrease in downtime

  • Improved resource utilization, with 60% lower server requirements

Additional Case Studies