Ahamove Migrates 1,000 Containers to Amazon EKS to Improve App Stability

Vietnam’s delivery sector presents a vast opportunity as domestic income levels rise and ecommerce becomes widespread. But the country’s logistics industry, including last-mile delivery, continues to face high costs, 6–12 percent higher than similar costs in Thailand, China, and Malaysia. Widespread fragmentation in Vietnam’s logistics market has also created a bottleneck preventing the country from tapping its full commercial potential.  

Ahamove is an on-demand logistics service founded in 2016 with a mission to address market inefficiencies and high costs for consumers. The Ahamove app connects its customers, mainly small businesses, with 50,000 professional drivers and their vehicles in its network. As of late 2021, the app had 300,000 monthly users.

Since launching, Ahamove had been following a multi-cloud strategy. In 2019, it began containerizing workloads using Kubernetes and a cloud-based container orchestration service. However, the business faced limitations on scaling and customization inherent to the cloud service it was using for container workloads. Its engineers wanted more granular, flexible control over container clusters to manage costs and adjust to evolving market dynamics. 

In 2021, Ahamove ran a proof of concept for its microservices infrastructure on Amazon Elastic Kubernetes Service (Amazon EKS) and found the service easier to work with. The ability to tailor nodes in Amazon EKS, customized to Ahamove’s business needs and service volumes was valuable to Ahamove. “By running our containers using Amazon EKS, we can choose nodes with different AMIs [Amazon Machine Images] and varying costs to match the strategic needs of the company,” says Ngon Pham, chief executive officer at Ahamove. “Controlling costs on the backend helps us keep costs low for our customers.” 

By then, the business was running most of its non-containerized applications on Amazon Web Services (AWS). “We decided to centralize as much of our technology as possible on the AWS Cloud,” Pham says. “AWS does a great job of integrating native compute, storage, and data services with automated third-party tools, and there’s a lot of support from the AWS team as well as the wider engineering community.” 

As of January 2022, Ahamove is managing nearly 1,000 containers on Amazon EKS. It’s using Amazon Elastic Compute Cloud (Amazon EC2) Spot Instances in conjunction with AWS Savings Plans to manage costs, and has regular cost optimization meetings with AWS. 

In addition to lower costs, scalability and reliability have improved since migrating to Amazon EKS. Uptime on the Ahamove app is 99 percent or higher, and engineers no longer spend time debugging or addressing scaling-related issues. Instead, they’re exploring more features on the AWS platform, and Ahamove has signed its team up for AWS Certification courses. Pham says, “Stability is guaranteed on AWS, and we’ve greatly reduced our infrastructure risk since migrating to Amazon EKS.”

Ahamove is also using AWS data and analytics services to enhance its customer experience. The business has built a data pipeline to track GPS locations for its drivers and to communicate delivery updates to customers in real-time. Ahamove uses Amazon Simple Storage Service (Amazon S3) as a data lake, and Amazon Athena with AWS Lambda to perform computations and analytics. Pham concludes, “AWS offers a variety of data services at reasonable prices that facilitate real-time updates for data streams to support our analytics use cases.”