amaysim Saves 75% on Compute Costs by Running Amazon EC2 Spot Instances

Mobile virtual network operator amaysim wanted to optimize its cloud architecture and compute costs after completing its migration to Amazon Web Services (AWS) in 2017. Originally, the company lifted and shifted all its infrastructure to Amazon Elastic Compute Cloud (Amazon EC2), which provides secure, resizable compute capacity in the cloud.

While completing its cloud migration, amaysim continued to improve the resilience of its compute infrastructure to help it handle unpredictable spikes in traffic and minimize interruptions in service. In 2020, amaysim switched to using Amazon EC2 Spot Instances—which let users take advantage of unused Amazon EC2 capacity and are available at up to a 90 percent discount compared to Amazon EC2 On-Demand Pricing—for its baseload capacity. As a result, amaysim was able to reduce its compute costs by 75 percent and improve the performance and availability of its online services.

Based in Australia, amaysim has provided no-commitment mobile phone plans to over 1.2 million customers since 2010. The company prides itself on its customer-centric approach, and in 2021, it received the Mobile Provider of the Year award from WhistleOut.

When the company was first established, amaysim was leasing on-premises data centers in Sydney and Brisbane for its compute infrastructure. However, the company had difficulties scaling to meet demand as it attracted more business. In 2014, amaysim and some of its partners faced significant scaling and operational challenges, which prompted the company to begin the migration of its infrastructure to AWS.

Consequently, amaysim wanted to find new ways to save on compute costs so that it could reallocate its funds and improve the user experience. “Our brand differentiators are that we don’t have contracts and that we make things simple for our customers,” says Peter James, IT operations director at amaysim. “Key for us are low capital investment and delivering the best possible experience for our customers.”

To become a cloud-first company, amaysim started experimenting with different instance types that would meet its licensing and compute needs. In 2015, amaysim’s teams created Auto Scaling groups—which contain collections of Amazon EC2 instances that are treated as logical groupings for the purposes of automatic scaling and management—and combined different Amazon EC2 instances. Initially, amaysim purchased Amazon EC2 Reserved Instances, which provide a significant discount (up to 72 percent) compared to On-Demand Pricing. Using Reserved Instances, the company was able to support continuous integration and delivery. “We are trying to achieve continuous integration and delivery with our infrastructure as well as our applications,” says Isaac Gittins, cloud architect at amaysim. “We programmed our deployment pipelines to choose different instance classes if required.” The company also adopted AWS CloudFormation, which lets users model, provision, and manage AWS and third-party resources by treating infrastructure as code. By moving to this model, amaysim was able to eliminate manual deployment tasks and streamline its software delivery workflows.

Next, amaysim adopted dual Auto Scaling groups, using a combination of Reserved Instances and Spot Instances. In 2017, the company completed its full cloud migration, and after it received a spike in customer demand, its Spot Instance usage grew from 20 percent to 44 percent. As of 2021, amaysim runs its workloads across six instance families, and all of its production workloads run 100 percent on Spot Instances. “We’ve improved the resilience of our application and reduced the chance of outages using diversified Spot Instances,” says Gittins. By diversifying its instance types, amaysim has achieved greater scalability and saved 75 percent on compute costs.

The company has also enhanced the performance of its billing system and self-service portal by using Spot Instances. Previously, intensive batch requests could take up to 7 minutes to process on its billing application. However, amaysim has set up its system to split a large request into smaller requests, accelerating the time it takes to run a transaction. Now, the company can perform intensive batch requests in 10–12 seconds without negatively impacting the customer experience. “Having a more resilient application makes everybody happy,” says Gittins. “Our website is our first touchstone, and we have fewer customer-facing errors.” Additionally, since adopting Spot Instances, amaysim’s development team must complete fewer tasks related to maintaining and managing the billing system and self-service portal from the backend.

Moreover, amaysim has containerized its workloads for its customer-facing applications, microservices, and scripts using Amazon Elastic Container Service (Amazon ECS), a fully managed container orchestration service that makes it easy for users to deploy, manage, and scale containerized applications. By containerizing its workloads, the company has benefited from improved reliability and performance of its website and applications. Since making these changes, amaysim has also improved its uptime and maintained significant levels of customer satisfaction. For instance, amaysim has sustained a high Net Promoter Score, a measure of customer experience and predictor of business growth. The average score for cellular phone service companies is around 30—amaysim’s Net Promoter Score is nearly double the average score. 

The company plans to continue innovating and optimizing compute costs. Currently, amaysim is exploring the possibility of using Fargate Spot Pricing for Amazon ECS, which lets customers run interruption-tolerant Amazon ECS tasks on spare capacity at up to a 70 percent discount off the regular AWS Fargate price.

The company is also looking at implementing guardrails using AWS cloud solutions, including AWS Control Tower, which provides an easy way to set up and govern a secure, multiaccount AWS environment. By having guardrails in place, amaysim hopes to promote greater uniformity across its solutions and increase operational efficiency. “We are always looking for new ways to improve our business,” says James. “We’re able to remove barriers to innovation using AWS.”