

Foot Locker cuts AWS Infra Costs by 50% on Adobe Commerce (Magento) with Codilar

Footwear
Company Brief
Foot Locker is a global footwear and athletic apparel retailer with thousands of stores and a high-traffic Adobe Commerce (Magento) eCommerce platform. The brand manages a large digital catalog and supports millions of daily customer interactions across its online storefronts.
In Indonesia, the brand’s digital operations are managed by PT MAP FTL Adiperkasa, part of the MAP Group, Southeast Asia’s largest lifestyle retail group. The platform supports approximately 80,000 products across six websites and handles nearly 6 million daily requests from customers browsing, searching, and purchasing online.





Infrastructure Stability
Cloud Cost Optimization
Database Resilience
Performance Optimization
Security Hardening
Application and Hardware Compatibility
Project Overview
Codilar partnered with Foot Locker to stabilise and optimise its high-traffic Adobe Commerce (Magento) infrastructure running on AWS. The engagement focused on identifying architectural risks, reducing infrastructure costs, and improving platform resilience through a comprehensive audit-first approach. By right-sizing compute resources, strengthening database reliability, and refining system configurations, Codilar helped build a more efficient and high-performance foundation capable of supporting millions of daily requests.
Business Challenges
Over-provisioned EC2 Auto Scaling configuration increased baseline infrastructure costs
Expiring AWS Savings Plan risked 20-30% higher compute expenses
Single RDS database instance created a critical single point of failure
PHP-FPM misconfigurations generated excessive request load on compute nodes
Inefficient microservices configuration added unnecessary request overhead
Suboptimal Redis and Varnish caching reduced traffic absorption efficiency
Outdated cron jobs and background services consumed infrastructure resources
Infrastructure security required stronger endpoint and vulnerability protection
Codilar's Solution Approach
Migrated EC2 instances from C5 to M7i for improved price-performance
Right-sized Auto Scaling from Min 40/Max 60 to Min 5/Max 40 based on traffic patterns
Upgraded instance capacity from 2xlarge to 4xlarge while reducing active nodes
Migrated MySQL to MariaDB and implemented a multi-node cluster with read replicas
Fixed PHP-FPM and microservice misconfigurations causing repeated 503 errors
Optimized Varnish and Redis caching to improve traffic handling and response times
Integrated CrowdStrike and Qualys security alongside existing Fastly CDN and WAF
Your One-stop
Total Commerce
Partner
We engineer high-converting eCommerce webstores that scale faster.
Book a 30min Consultation CallWhen millions of customers depend on your platform daily, 'good enough' is never good enough.

Jameel Ahmad Ansari
Head of Engineering, Codilar
Behind the Platform That Handles ~ Six Million Requests a Day
When Premium Experience Meets Bonafied Complexity
For most fast-growing digital commerce platforms, success brings a new kind of challenge. The more customers arrive, the more pressure the underlying infrastructure must absorb. Foot Locker’s eCommerce ecosystem was already operating at an impressive scale. Millions of requests flowed through the platform every day as sneaker enthusiasts explored collections, checked availability, and completed purchases across multiple storefronts.
From the outside, everything appeared to be running efficiently. But when Codilar began its pre-engagement infrastructure audit, a different story started to emerge behind the scenes. The system was carrying inefficiencies and architectural risks that had accumulated over time. None of them had caused a major failure yet, but together they were increasing operational cost, limiting efficiency, and introducing potential points of disruption.
Recognizing this, the MAP Group team wanted more than routine infrastructure maintenance. They wanted clarity, stability, and a roadmap that would ensure the platform could grow confidently with demand.
Right-Sizing the Engine Behind the Platform
One of the first discoveries lay in the platform’s compute layer. The infrastructure supporting Foot Locker Indonesia’s websites was configured to run a minimum of 40 EC2 instances at all times. While this guaranteed capacity, it also meant the system was constantly operating at a scale that didn’t always reflect real traffic patterns.
In simple terms, the platform had horsepower to spare, but it was paying for it every minute of the day. Codilar analysed real traffic behaviour across the environment and redesigned the Auto Scaling configuration to better align with actual demand. By carefully recalibrating the scaling thresholds, the team was able to reduce the minimum compute baseline while preserving peak capacity for high-traffic events. The result was a leaner infrastructure engine that remained powerful when needed but significantly more efficient during normal operations.
Removing the Platform’s Biggest Hidden Risk
While compute efficiency presented an opportunity, the database architecture revealed a far more critical concern. At the center of the entire platform sat a single large RDS database instance. Every storefront, product search, and checkout interaction depended on that single node. If it failed, the entire digital operation could go offline.
For a platform handling millions of daily requests, that kind of dependency represented a serious vulnerability. Codilar redesigned the database architecture to introduce a multi-node MariaDB setup with read replicas, distributing the workload, and introducing fault tolerance into the system. With this shift, the platform moved away from a fragile single-point architecture to one designed for reliability and resilience.

We Make An Impact
Through Our Work
AWS Infrastructure
Cost Reduction
Reduction in Minimum
EC2 Instance Baseline
Daily Requests Supported
Across Multi-Site Infrastructure
Products Supported
in Catalog
503 Errors After
Infrastructure Optimization

We Make An Impact
Through Our Work
AWS Infrastructure
Cost Reduction
Reduction in Minimum
EC2 Instance Baseline
Daily Requests Supported
Across Multi-Site Infrastructure
Products Supported
in Catalog
503 Errors After
Infrastructure Optimization

Fine-Tuning the Infrastructure Ecosystem
Beyond the major architectural changes, the audit also revealed several configuration-level inefficiencies quietly affecting system performance. Some PHP-FPM configurations were generating more requests than necessary. Certain microservice interactions were adding extra load across the application stack. Even outdated background jobs were continuing to consume resources long after their usefulness had passed.
Individually, these issues seemed small. But across dozens of servers and millions of daily requests, their impact multiplied. Codilar systematically addressed each of these areas. Request handling was optimized, caching layers were strengthened, and unnecessary background processes were retired. The improvements may not have been visible to customers browsing the storefront, but they significantly increased the efficiency and stability of the platform supporting every interaction. Alongside these optimizations, Codilar also introduced production incident alert mechanisms and simplified deployment automation processes, enabling faster issue detection and more reliable, low-risk releases across the platform.
Strengthening Security at Every Layer
With infrastructure efficiency and resilience addressed, the next priority was strengthening the platform’s security posture. Foot Locker already leveraged Fastly’s CDN and edge WAF protection to filter and mitigate threats at the network edge. However, deeper visibility and protection were still required within the AWS infrastructure itself.
To close this gap, Codilar implemented CrowdStrike endpoint protection and Qualys vulnerability management, as suggested by the Foot Locker team, introducing continuous monitoring and threat detection directly at the infrastructure layer. This created a stronger defence-in-depth architecture, where both the edge and the origin infrastructure work together to detect, monitor, and mitigate security risks.
A Stronger Foundation for Scale
Codilar's partnership with Foot Locker Indonesia had quantifiable effects on infrastructure performance, cost effectiveness, and dependability. The platform currently operates on an AWS architecture that has been greatly simplified, resulting in a 50–60% decrease in infrastructure expenses and a reduction of the baseline EC2 instance count from 40 instances to roughly 5 nodes. The platform's most significant single-point-of-failure risk was removed at the same time via a revised database architecture, improving dependability for an ecosystem processing about 6 million requests every day. Additionally, the optimization removed persistent 503 errors, stabilizing the platform and drastically lowering the infrastructure needs.
With extensive knowledge of Adobe Commerce (Magento), Shopify, Shopify Plus, Pimcore, Akinon, and other modern commerce platforms, Codilar assists international brands in reducing infrastructure complexity, safeguarding performance, and creating digital commerce platforms that can grow with assurance.



