Delivering Cost Savings and Performance Improvements on AWS

Home / Resources / Case Study / Delivering Cost Savings and Performance Improvements on AWS
Sourced case study Delivering Cost Savings and Performance Improvements on AWS

About the Client

A leading provider of digital content protection for the film, entertainment, and broadcast industries was looking to mature its cloud management strategy. Sourced Group (Sourced) an Amdocs company investigated three of the company’s AWS deployments to identify areas for improvement.

Outcomes included:

  • Detailed analysis of performance and cost management opportunities across 89 instances.
  • Targeted and practical recommendations to deliver performance improvement in the cloud.
  • Expert guidance on how to handle potential cost implications of suggested changes.

Challenge: A Complex Cost and Performance Management Scenario

With 89 active instances of different sizes held across multiple time zones and AWS regions, cloud management was a significant challenge for this cybersecurity specialist. It needed to look at how instances were being used then identify ways to improve budget control and application performance. As the company’s cloud managed services partner, Sourced handled this activity as part of the ongoing cloud maturation journey.

Solution: A Methodical Approach Reveals Areas for Improvement  

The first step was to categorise the instances. They were all Amazon EC2 T2 instances, a low-cost, general purpose instance type which provides a baseline level of CPU performance with the ability to ‘burst’ above this when needed. As one of the lowest cost Amazon EC2 instance types, they are a good and straightforward choice for general-purpose applications, especially for companies that are new to the cloud.

We identified that most of the instances were the T2.small classification. From a cost perspective, they were ideal for the company’s non-CPU heavy applications, rarely bursting over the limit. However, it was a different story for memory utilisation where up to 60% of the instances were running at over 90% of memory. This can lead to application failure, with EC2 instances failing status checks due to insufficient memory availability for additional processes.

Simply increasing the T2 instance size would naturally incur higher costs (in this case it would have potentially doubled monthly cloud spend). So, we outlined two alternative recommendations for the client to consider:

  • Option 1 – Switch from T2 instances to the Amazon EC2 T3 series. T3 instances provide double the CPU credit allocation of T2, bringing a 30% improvement in price to performance ratio. This would deliver better stability and reliability for applications requiring higher CPU performance while reducing overall costs.
  • Option 2 – Containerise the applications and use multi-regional EKS clusters or AWS Fargate to run them. This makes everything cheaper and easier to manage. It also has the added advantage of facilitating deployment to other cloud providers if that’s required as part of a multi-cloud strategy.

Outcome: Enabling Progress on the Cloud Maturity Journey

Detailed analysis of the instances enabled Sourced to make practical recommendations for performance improvement that also satisfied cost management needs. These insights will empower the company’s in-house team to make well-informed decisions as they make strides towards cloud maturity.

This activity was part of a wider managed services partnership enabling the cybersecurity specialist to enhance its cloud-based operations.