Ferns N Petals Group has been a leading name in the gifting industry for over 25 years, helping customers celebrate special moments with fresh flowers and thoughtful gifts. Founded in 1994 with a focus on flower bouquet delivery, the company has since expanded into multiple verticals, including retail, e-commerce across India, UAE, Singapore, and Qatar, weddings and events, luxury venues, handicrafts, and media. Today, Ferns N Petals continues to redefine celebrations with its diverse range of services and global presence.
Challenges
Solutions
FnP running all workloads in a single Amazon EKS cluster with separate node pools for stateless applications and stateful applications (such as databases). While the setup functioned well during peak hours, it lacked efficient scaling during idle hours, leading to unnecessary costs. So FnP looking for a solution to optimize performance and reduce operational costs through better scaling strategies.
To address these challenges, Brio proposed a multi-cluster architecture in Google Kubernetes Engine (GKE) and separated stateless applications and stateful applications into distinct clusters, allowing for independent scaling and improved resource allocation. Additionally, Brio implemented GKE Multi-Cluster Gateway API, enabling seamless service-to-service communication across clusters while maintaining security and performance. This architecture provided better cost control, optimized resource utilization, and improved scalability.
Business Impact
The migration to GKE with a multi-cluster architecture brought significant improvements to the customer's cloud operations. Separating stateless and stateful workloads into dedicated clusters enhanced performance and reliability, reducing latency and ensuring better resource utilization. The implementation of GKE Multi-Cluster Gateway API streamlined cross-cluster communication, making traffic management seamless and improving operational efficiency. Additionally, the new architecture enabled dynamic scaling based on real-time demand, allowing the system to efficiently handle peak loads while minimizing costs during idle hours. With structured cluster separation, resource allocation became more predictable, reducing operational overhead and simplifying management. Overall, the migration provided a cost-effective, scalable, and high-performing cloud environment that aligned with the customer's business needs.
