Title

AWS re:Invent 2023 - Reducing TCO and downtime by migrating to AWS Fargate (CON322)

Summary

Jeremiah Dooley and Praveen Edam presented on reducing Total Cost of Ownership (TCO) and downtime by migrating to AWS Fargate.
They discussed the importance of defining TCO, measuring it, and tracking it over time.
The session highlighted the benefits of adopting a serverless model, particularly in terms of infrastructure management, security, and cost savings.
AWS Fargate was presented as a key component of the serverless ecosystem, offering a managed container orchestrator that simplifies operations and reduces costs.
The speakers shared customer success stories, demonstrating significant cost reductions and operational improvements after adopting AWS Fargate.
A phased migration journey was outlined, starting from on-premises environments to EC2, then to ECS with EC2 launch mode, and finally to ECS with Fargate launch mode.
The talk concluded with insights on dynamic scaling, cost modeling, and the potential of serverless functions like AWS Lambda and AWS App Runner for further modernization and TCO reduction.

Serverless Model Adoption: The serverless model is increasingly popular among AWS customers due to its ability to offload infrastructure management to AWS, leading to lower TCO and improved operational resilience.
ECS and Fargate: Over 65% of new Amazon container customers use ECS, and among them, over 70% run their workloads on Fargate, indicating a strong trend towards serverless compute engines.
Customer Success Stories: Real-world examples from highly regulated industries, such as financial services, showed that migrating to AWS Fargate can lead to substantial cost savings (up to 35% in some cases) and operational benefits.
Phased Migration Approach: A step-by-step migration approach is recommended, starting with a lift-and-shift to EC2, then containerization with ECS, and finally adopting Fargate for a serverless environment.
Dynamic Scaling and Cost Modeling: AWS offers tools like Compute Optimizer and Fargate Spot, which help customers optimize their resource usage and costs based on traffic patterns and usage.
Serverless Functions: AWS Lambda and AWS App Runner are suggested for further modernization, offering a pay-per-request model and a fully managed experience, respectively, which can lead to additional TCO reductions.