Performance Testing and Scalability Optimization: Scaling for Success
Optimizing Infrastructure, Testing, and Monitoring for Peak Performance
The Opportunity
A leading global travel and tourism operator needed a complete digital transformation of its IT infrastructure and in a previous collaborative project, SourceFuse recommended that the customer migrate and modernize their Microsoft workloads on AWS, offering all the benefits that come with moving from on-premise to the cloud.
As part of SourceFuse’s ongoing support, the subsequent request regarding performance optimization was to ensure that all the application components would be capable of rapidly scaling up or down based on demand.
Key Challenges
The key challenges the customer was looking to overcome were as follows:
- Architecture Complexity: the customer had a highly complex application architecture with interdependent modules of legacy on-prem services, migrated cloud components, and AWS managed services.
- Data Volume Spikes: varying volumes across the modules, due to specific data splits for unique identifiers, were requiring different scaling methods for each component. Frequent spikes also occurred as a result of having vendors who make daily rather than monthly or yearly updates.
- Inability to Scale Configuration: there was varying CPU memory utilization at component level due to existing client code.
- Lack of End-to-End Visibility: this was as a result of the architecture being a mix of legacy on-prem and AWS cloud infrastructure.
The Solution
In order to address the challenges above, SourceFuse carried out a comprehensive assessment of the customer’s IT infrastructure. In doing so, it was able to identify services where component level testing could be conducted in isolation for early defect identification, leveraging AWS SDK integrations to simulate traffic. In addition, SourceFuse performed end-to-end testing for validating data integrity for components that could not be tested in isolation.
To further help address the customer’s key challenges, SourceFuse proposed implementing different scaling criteria, like CPU and memory utilization or Amazon SQS (message queuing service) for each component based on extensive testing to mitigate the spikes.
As AWS Premier Partners, SourceFuse was also able to maximize the use of advanced AWS services in order to test the limits of AWS Managed Services. It was also able to introduce third-party service DataDog for real-time monitoring of the end-to-end application flow.
The Results
As a result of the performance testing and recommendations, SourceFuse helped the customer to achieve:
- Component and integrated testing for application components and AWS Managed Services under projected growth volumes (up to 1.2M RPM for DataLake).
- Validation of the implemented scaling criteria, such as CPU and memory utilization or SQS Message Queue for each component, based on extensive testing to ensure minimal errors on scaling.
- Introduction of DataDog monitoring dashboards to give an end-to-end view of the application for both cloud and legacy systems.
- Additional dashboards and alerts were set up in AWS CloudWatch to optimize the tracking of usage limits in AWS Managed Services.
About The Customer
As one of the top travel technology companies, our customer empowers every segment of the travel and hospitality industry with real-time intelligence, connectivity to the largest demand and supply network, as well as targeted digital marketing solutions.
Trusted by multiple Fortune 500 Travel and Hospitality companies in the world, they offer a comprehensive suite of SaaS based products designed to enable their customer base to become digital-future & market ready.
