How We Saved Big Using AWS Cloud Financial Management Guidelines?

Introduction

I recently worked on a project migrating Windows servers from private datacenter and Azure, into AWS Cloud. The major goal of the project was to save on annual costs incurred by the client after migration. The project was highly successful from the viewpoint of meeting business requirements and maximizing business value at the lowest possible cost. We spend maximum amount of time in cost savings activities for the lowest possible Total Cost of Ownership (TCO) covering a period of three years. In this article, I will share the lessons learnt so that it helps organizations to enable a cost- conscious culture that drives accountability across all teams to drive expected business outcomes. We used AWS Cloud Financial Management guidelines to fine tune every possible knob for cost savings.

What is Cloud Financial Management?

Cloud Financial Management is a set of activities that enables businesses to manage, optimize, and plan costs as they run workloads in cloud. It helps to find the right balance between agility and control, while improving per unit economics and efficiencies. Effective CFM requires a partnership between finance and technology teams, and spans activities across the following four pillars:

• See (Measurement & Accountability)
• Save (Cost Optimization)
• Plan (Planning & Forecasting)
• Run (Cloud Financial Operations)

See: Measurement & Accountability

• Provisioned organization account structure using AWS Control Tower and implement tagging dictionary for mapping cost and usage for workloads in each account
• For proactively detecting and addressing cost variances, created AWS Budgets to track and take actions. AWS Budgets provides a Cost Explorer graph to help see incurred costs and usage
• For driving cost aware cloud consumption, showed charges for each business units using AWS Cost Explorer reports
• For validating decisions and outcomes, measured and circulated efficiency and review it regularly

Save: Cost Optimization

• Prevent resource waste using instance scheduler for non-production environments
• Manage resources in a cost-efficient manner by rightsizing resources. Select the right size of the resources by considering each usecase/application
• Apply the most efficient cost savings plan for all the resources across all accounts
• Match capacity with demand by exploiting Auto Scaling capabilities across a number of AWS services to enable workloads scale when needed

Plan: Planning & Forecasting

• Understanding cost, especially around dedicated host usage and network traffic was the key to planning and forecasting costs
• Strategic business planning around licensing cost is a must for understanding the current and forecasted cost
• To make cost based decisions faster, implemented proof of concept
• Implemented interactive, custom dashboard for exploring cost, used by the entire project team for planning and forecasting

Run: Cloud Financial Operations

• Treated Cloud Financial Management as a first-order initiative alongside others, thereby increasing the potential for improving cloud consumption efficiencies.
• Invested in people (training), governance (process), and tools (automation) as the workloads scale.
• Eliminate external third party costs by leveraging AWS managed services. e.g. we used AWS KMS for PKI and AWS ACM for public/private certificates

Optimize Compute For Cost Savings

• Since we migrated windows servers, leveraging the current licensing terms was the key in savings cost. We used BYOL option in AWS Application Migration service and managed licenses using AWS License Manager
• AWS Instance Scheduler is a simple AWS provided solution that enabled us to easily configure custom start and stop schedules for Amazon EC2 and Amazon RDS instances. The solution was easy to deploy and helped reduce operational costs for development and production environments.
• Right sizing the instances and re-reviewing the current consumption helped to achieve better price-to-performance ratio
• Automated compliance reporting using AWS SSM patch baselines helped to achieve consistency across all accounts. Also, it helped to reduce costs incurred by third party products
• Detailed comparison between AWS pricing for reserved instances/savings plan/compute savings plan for maximizing the business value from infrastructure costs

Optimize Network For Cost Savings

• Networking was based on AWS Direct Connect, AWS Transit Gateway, and AWS Site-to-Site VPN
• Considered data transfer options/pricing for data movement between availability zones and regions
• Considered data transfer between VPCs for optimizing networking costs
• Created comprehensive security using AWS WAF and eliminated using similar solutions from the marketplace

Optimized Storage For Cost Savings

• Deleted any EBS volumes not in use
• Implemented life cycle policies for S3 storage classes
• Moved AWS Backup to S3 Glacier for significant cost savings
• Involved stakeholders to define backup plans and retention periods

Conclusion

Cost optimization is an ongoing activity with every team member contributing, to achieve the maximum business value at lowest possible cost. We worked closely with stakeholders, finance and technology teams for continuously measuring cost and usage. We spend significant amount of time for finalizing strategies to determine and predict costs associated with current and future cloud workloads. Focusing on the design principles of Cost Optimization pillar from AWS Well Architected framework, helped us to realize the economic benefits of the cloud, quite accurately. Needless to say, It was the most important factor for the client!