Our client, a developer of scalable and secure eCommerce platforms, introduced an enterprise-level solution designed to handle large catalogs and high transaction volumes. To ensure the platform’s robustness and reliability, our task was to set up a suitable infrastructure and conduct comprehensive load testing. The aim was to create a resilient system capable of auto-scaling to manage varying loads, thus ensuring optimal performance and stability under high-demand scenarios.

2 weeks

2 DevOps Engineers

eCommerce

IT

The challenge was to build a system that could handle significant traffic spikes while maintaining high performance and reliability. This required a carefully planned infrastructure setup with separate groups of servers for MySQL and Redis databases, as well as auto-scalable servers for PHP and ElasticSearch/OpenSearch. The infrastructure needed to ensure seamless scaling up during peak loads and scaling down during low-demand periods to optimize resource usage and reduce costs.

To address the challenge of managing significant traffic spikes while maintaining high performance and reliability, we designed a carefully structured infrastructure. The solution involved the configuration of four core servers for the primary infrastructure:

These servers formed the backbone of the system, ensuring smooth operation under high traffic conditions. In addition to the primary infrastructure, auxiliary servers were used to handle internal processes like log management and incident response. While the project could run without these auxiliary servers, their inclusion ensured the integrity of log management and proactive incident response. Alternatively, clients could opt for cloud-based solutions for log handling and issue forecasting.

To meet the demand for dynamic scalability, we implemented separate groups of servers for MySQL and Redis databases, along with auto-scalable servers for PHP and ElasticSearch/OpenSearch. This allowed the system to scale up during peak traffic periods and scale down during low-traffic times, optimizing resource usage and reducing operational costs.

We selected Kubernetes as the core of the infrastructure setup due to its strong scaling capabilities. The system was designed for seamless auto-scaling, deploying additional resources during high demand and scaling back when the load decreased, ensuring stability and cost-efficiency.

For enhanced performance, we integrated Varnish caching, which significantly reduced response times. Throughout the load testing phase, we used advanced monitoring and profiling tools to identify and address performance bottlenecks in the application, database, and caching layers.

During load testing, the system demonstrated strong performance across all components. The tests evaluated how well the infrastructure managed traffic, load balancing, and the interaction between the application (platform core and add-ons), application servers, databases, and caching layers.

The system maintained high performance until reaching 82 Requests Per Second (RPS), where performance degradation began. This threshold corresponded to approximately 1,300-2,000 concurrent virtual users (as monitored by Google Analytics). These results confirmed the infrastructure’s ability to handle significant loads efficiently, offering scalability and reliability even under intense traffic conditions

The infrastructure solution, combined with extensive monitoring and profiling, enabled us to identify and resolve several performance bottlenecks, which were subsequently addressed by the client’s team. As a result, the optimized version of the enterprise platform now delivers significantly faster response times and improved performance stability. Our infrastructure design allows the application to efficiently process up to 15,000,000 products with advanced search filters in just 20-90 milliseconds, surpassing current eCommerce performance standards.

By clicking Submit, you agree with Privacy Policy

Ready to elevate your online presence with Scalesta hosting solutions?