Title

AWS re:Invent 2023 - Top cost optimization recommendations for Microsoft workloads on AWS (ENT211)

Summary

• The session focused on cost optimization strategies for Microsoft workloads on AWS.
• Bill Pfeiffer introduced the Microsoft on AWS Cost Optimization Initiative (MAKO), which offers 44 recommendations across seven Microsoft domains.
• The MAKO initiative is accessible via the AWS site and includes a guide with technical guidance, effectiveness, and cost tracking.
• The AWS Optimization and Licensing Assessment (OLA) helps build migration and licensing strategies.
• Right-sizing and choosing the right EC2 instance type can lead to significant cost savings.
• AWS Compute Optimizer automates right-sizing and provides recommendations for EC2 instances, ECS Fargate tasks, EBS volumes, and Lambda functions.
• Dedicated hosts and Bring Your Own License (BYOL) can reduce costs for software without license mobility, like Windows Server.
• SQL Server cost can be reduced by using Software Assurance, downgrading editions, and using Developer Edition for non-production environments.
• Migrating from GP2 to GP3 EBS volumes can save costs and improve performance.
• Savings plans and reserved instances offer discounts for committed usage.
• Modernizing.NET to newer versions and using AWS Graviton can lead to cost and energy savings.
• The AWS Carbon Footprint tool helps track the environmental impact of AWS workloads.
• The session highlighted the intersection of cost optimization and sustainability, emphasizing tools like EC2 Spot Instances, autoscaling, and AWS Lambda.
• Attendees were encouraged to engage with the MAKO team for assistance and to use the provided resources and contact information.

Insights

• The MAKO initiative is a significant effort by AWS to address the specific cost optimization needs of Microsoft workloads, acknowledging the unique challenges posed by licensing and enterprise applications.
• The OLA program is a critical first step for organizations to understand their current licensing situation and potential cost savings before migrating or optimizing Microsoft workloads on AWS.
• AWS Compute Optimizer is a key tool for organizations to automate the right-sizing of their instances, potentially unlocking substantial cost savings.
• Dedicated hosts are essential for complying with Microsoft's licensing restrictions and can offer significant cost reductions when used with BYOL strategies.
• SQL Server licensing costs can be a major factor in the total cost of ownership, and strategies like using Software Assurance, downgrading editions, and leveraging Developer Edition for non-production can lead to large savings.
• Migrating to GP3 EBS volumes is an easy and impactful way to reduce costs and improve performance for storage-intensive workloads.
• Savings plans and reserved instances are effective cost-saving mechanisms for predictable, steady-state workloads.
• Modernizing.NET applications to run on Linux and Graviton processors not only reduces licensing costs but also improves performance and sustainability.
• The AWS Carbon Footprint tool and other sustainability-focused tools demonstrate AWS's commitment to helping customers achieve their environmental goals alongside cost optimization.
• Engaging with the MAKO team and utilizing the resources provided can help organizations navigate the complex landscape of cost optimization for Microsoft workloads on AWS.