Scalable Cloud Architecture Design for Fluctuating Traffic on AWS

Project Overview

Duration

3 Months

Services

Technologies

SNS/SQS
DynamoDB
ECS
AWS Lambda

Our client (under NDA) approached us with a challenge: designing and building robust cloud architecture to address fluctuating traffic patterns on their insurance platform from one of their integrations. They specifically needed a cost-effective solution that would maintain high performance while in line with rigorous SLA requirements.

Challenge

In order to ensure the project’s success, we needed to design a proper cloud architecture. This posed several challenges we needed to address:

Low initial transaction volume: From the outset, we didn’t expect much traffic on the platform. But we did anticipate future surges driven by the hefty marketing investments of our client’s partner. The challenge at hand? Being prepared for the traffic influx while efficiently managing costs during periods of low activity.
Scalability requirements: With the anticipated growth of the distribution channel, the architecture needed to be highly scalable. This was crucial to handle the expected increase in traffic and ensure seamless operations during peak periods.
Fluctuating traffic patterns: Insurance sales tend to spike in the morning and afternoon, leaving lulls in between, which presented a unique challenge. The architecture needed to be capable of handling these fluctuations efficiently, ensuring optimal performance during peak hours and cost-effectiveness during off-peak periods.
Response time SLA: Meeting service-level agreements (SLAs) for response times was critical. The system needed to consistently deliver fast response times, regardless of the volume of requests, to maintain a positive user experience.
Batch Data Exchange: Another challenge was the need for batch data exchange once a month. Efficiently managing large volumes of data within specified timeframes without disruptions to ongoing operations was a puzzle to solve.

Our Approach

The first step in addressing the challenges faced in designing a resilient cloud architecture involved selecting the right serverless technologies. Given the sporadic nature of traffic, we opted for cloud-native AWS technologies to optimise fixed costs. Leveraging tools such as

SNS/SQS, DynamoDB, ECS, and AWS Lambda allowed us to efficiently handle resources and scale up or down as needed, ensuring cost-effectiveness without compromising performance, even during fluctuating traffic patterns. We carried out thorough performance testing to evaluate the system’s efficiency and scalability.

With maintaining a seamless user experience in mind, we decided to reduce the number of synchronous operations wherever possible. Thanks to implementing an asynchronous architecture, we were able to process spikes in traffic in the background without affecting the customer experience. This approach ensured that the system stayed both responsive and resilient, regardless of fluctuations in traffic volumes.
What’s worth pointing out is that our setup embraced event-driven architecture, making communication between system parts seamless and boosting overall responsiveness.

Effects

Thanks to the solutions we implemented, we successfully designed and built a robust cloud architecture that met the requirements of the project:

Scalability and performance compliance: the system is scalable and checks all the performance boxes, effectively addressing SLA demands for response time and reliability regardless of traffic volume. This ensures a seamless user experience under varying conditions.
Cost efficiency: infrastructure costs are significantly reduced during periods of low traffic, increasing only during peak hours. This approach demonstrates adaptive resource allocation, optimising cost-effectiveness.

Turn your project into a success story like this!

Want to unlock the success of your project? We hold the key. And it’s yours to take.

let’s talk