BigBasket Modernizes Kubernetes Traffic Management with Traefik's API Gateway

Overview

BigBasket, India’s largest online grocer, faced growing challenges in managing traffic across their Kubernetes infrastructure. As their e-commerce platform scaled, their microservices infrastructure became more complex and increasingly difficult to manage, leading to concerns over scalability and operational efficiency. Sushant Gulati, Senior Engineering Manager for Cloud, DevOps, and Platform, alongside Sayantan Bhattacharjee, Principal DevOps Engineer, led the effort to find  a robust API Gateway solution to simplify microservice traffic management, improve day-2 operations, and future-proof their infrastructure.

Challenge

Before Traefik, BigBasket’s network architecture relied on NodePort services and AWS ALB Target Groups (TG). While this worked at a smaller scale, this solution quickly reached its limits in terms of scalability and manageability. Each worker node was attached to every ALB target group, causing API throttling due to the excessive volume of register and deregister calls. As traffic grew, the system became unstable, eventually reaching AWS API limits. 

In addition to stability issues, scaling became difficult as microservices and hosts multiplied. Infrastructure management was increasingly time-consuming and complex. To address these challenges, they needed a solution capable of simplifying traffic routing while accommodating their growing needs.

Solution

After evaluating several API Gateway solutions, BigBasket’s team chose Traefik for its simplicity, scalability, and performance.

During their evaluation, BigBasket found that other solutions involved complex configuration and required numerous annotations to meet their custom needs, resulting in a high administrative burden. In contrast, Traefik’s use of Custom Resource Definitions (CRDs) made configuration much simpler and easier to manage and customize. Traefik provided a more intuitive solution that significantly reduced management effort while meeting their scalability and performance requirements.

Key factors in selecting Traefik included:

• Ease of Management: Traefik’s simpler configuration through CRDs allowed the team to manage their infrastructure with minimal effort, providing much-needed operational breathing room.
• Distributed Architecture: Traefik’s API Gateway offered a highly available, scalable, and secure architecture, making it a reliable choice for BigBasket.
• Advanced Traffic Management: Traefik’s API Gateway efficiently handled various routing scenarios, such as header-based, host-based, cookie-based, and path-based routing.
• High Performance: Traefik's performance, particularly its ability to handle increased traffic loads, was a critical requirement for BigBasket.

“With Traefik, deployment was a breeze, our success criteria were satisfied, and making this major change in our architecture was worry-free knowing we had enterprise support.” Sayantan Bhattacharjee, Principal DevOps Engineer at BigBasket 

Implementation

BigBasket deployed Traefik’s API Gateway as their primary solution across five Kubernetes clusters, all running in AWS. The new Traefik architecture became the foundation of their network, handling every transaction across their e-commerce platform. All their critical services are now routed through Traefik’s API Gateway, making it an indispensable component of their infrastructure.

The team significantly simplified their routing configuration by eliminating the need for multiple AWS target groups and application load balancers (ALBs). This not only reduced complexity but also allowed them to automate much of their setup and significantly improve the management of their architecture. 

Traefik’s API Gateway also enabled BigBasket to manage five different versions of the same service within a single cluster, supporting their testing needs without requiring additional tools or investments.

“Traefik is simple to configure, doesn’t have component sprawl, and has significantly reduced our service onboarding and route management efforts.” Sushant Gulati, Senior Engineering Manager – Cloud, DevOps and Platform at BigBasket 

Visibility into traffic management has improved substantially. With Traefik, the team gained deeper insights into traffic flow, making it easier to monitor and troubleshoot issues. The ability to track everything in real-time has been transformative and allowed them to efficiently manage and optimize their system. Everyone involved at BigBasket found Traefik simple to manage, resulting in smoother day-to-day operations.

Additionally, the support from Traefik Labs played a crucial role in their success. The team praised Traefik's proactive approach, noting how helpful their experts were during setup and in answering ongoing questions. Regular meetings with the Traefik Labs team ensured that BigBasket received continuous guidance and assistance in resolving any challenges.

“The support from Traefik has been very helpful, the team is very knowledgeable,  forthcoming, and open to discussions.” Sayantan Bhattacharjee, Principal DevOps Engineer at BigBasket 

Results

Since adopting Traefik’s API Gateway, BigBasket has experienced the following improvements:

• Scalability without limitations: Traefik has enabled BigBasket to scale seamlessly alongside their growing business The team has already increased their Traefik adoption by 50% within the first 12-months to support the expansion of their infrastructure needs. Had they remained on their previous solution, they would not have been able to scale to this extent.
• High availability and reliability: Over the past year and a half, BigBasket experienced no stability issues in their clusters, even during periods of significant traffic spikes, thanks to Traefik’s robust and stable environment.
• Simplified Traffic Management: Traefik eliminated the need for managing thousands of AWS Target Groups and ALBs, significantly reducing complexity and administrative overhead. This allowed BigBasket to handle a 2x increase in internal traffic without performance degradation.
• Reduced Complexity: By consolidating traffic management through Traefik’s API Gateway, BigBasket reduced the number of components needed to manage their system, freeing up engineering resources for more strategic tasks.
• Enhanced Visibility: Traefik’s built-in metrics and dashboard provided BigBasket with real-time insights into their traffic flow. This increased visibility, made troubleshooting easier, and allowed the team to respond quickly to system changes. The ability to monitor traffic at a granular level—down to requests-per-second (RPS)—has been critical in optimizing system performance and ensuring smooth operations.

“Traefik has exceeded our expectations; it has done a lot more for BigBasket than what is normally expected out of an API Gateway.” Sushant Gulati, Senior Engineering Manager – Cloud, DevOps and Platform at BigBasket 

Key Takeaways

Traefik’s API Gateway has transformed BigBasket’s approach to managing Kubernetes traffic,  delivering significant improvements in efficiency, scalability, and system stability. Traefik’s distributed architecture, automation, and performance have allowed BigBasket’s team to future-proof their infrastructure. Key outcomes include:

• 50% increase in Traefik usage, to support the expansion of their infrastructure.
• 2x increase in east-west traffic, handled seamlessly without impacting performance.
• Elimination of 7,000 NodePort configurations, drastically reducing the operational complexity.
• Mitigated AWS API throttling, resulting in fewer errors and operational headaches.

Sushant and his team at BigBasket continue to expand their use of Traefik, confident that it provides the flexibility, visibility, and stability needed to support their business at scale.

“Traefik has been a game-changer for us. Its ease of configuration, scalability, and seamless support have exceeded our expectations. Traefik is a strategic tool that has transformed how we manage traffic across our infrastructure. I would highly recommend Traefik to any organization looking to modernize their Kubernetes environment.” — Sushant Gulati, Senior Engineering Manager – Cloud, DevOps and Platform at BigBasket