Halaman ini belum diterjemahkan ke dalam bahasa Anda. Minta terjemahan

Implementation priorities

PDFRSS

The M&G Guide recommends that you implement your Cloud Financial Management capabilities with transparency in mind. This includes enabling your builder teams to see the financial impact of their cloud usage for the resources they provision, as well as to define specific controls related to the financial governance of your resources.

Enable Cloud Financial Management

Configure detailed information sources including Billing and Cost Management tools to create the reporting your organization needs. Regularly review (minimally on a monthly basis) the cost and usage by different dimensions to understand cost drivers. Establish organizational metrics, such as a unit metric to identify cost attribution categories as you scale. If required, ensure that your cost reporting includes all costs (labor, licensing, infrastructure, and more) to create the total cost of application management (TCAM).

Tag, track, and monitor resource costs across their lifecycle

A consistent and well-designed tagging strategy is required to manage and track costs across your AWS environments. Once resources in your environments are tagged, you must activate both AWS-generated tags and user-defined tags separately to use them in your cost reporting and analysis tools. Enforce tag options using distribution and preconfigured infrastructure as code templates for governance. Use tag policies to enforce and maintain consistent tags across your organization and resources.

Track resources over their lifetime and design your workloads to gracefully handle resource termination as you automatically identify and decommission non-critical or low utilization resources. Analyze the design, architecture, and all components of each workload or application for cost effectiveness, including license costs. Use Managed entitlements to track and help ensure that you have compliance with your established agreements while avoiding unexpected true-up bills for exceeding license limits. Determine if the component and resources will be running for extended periods (for commitment discounts), or dynamic and transiently running (for Spot or On-Demand Instances). Implement the appropriate pricing models for all components of your applications sourced from AWS Marketplace.

Establish mechanisms for cost governance

Create policies and mechanisms that define how resources are managed by your organization. The policies should cover cost aspects of resources and workloads, including creation, modification, and decommissioning over the resource lifetime. Create an obsolescence plan and defined retention period with lifecycle policies for resources as they are provisioned. Implement account structure, groups, and roles to help allocate costs and control who can create, modify, or decommission instances and resources in each group. Identify any new controls that could support a more efficient cost spend. Update your distribution of infrastructure as code templates in Service Catalog so that cost is transparent and only approved instance sizes are available in a self-service manner across your multi-account framework. Enforce tagging of resources as they are provisioned to ensure effective cost governance.

Continually optimize for cost efficiency

Review historic spend patterns to detect cost spikes (one-time or recurring) or continual cost increases, assuming 14–30 days of historical spend. Implement mechanisms to periodically identify and right-size instances based on current workload metrics and characteristics. This can be evaluated using AWS Cost Explorer, AWS Trusted Advisor, and AWS Compute Optimizer, along with AWS Partner tools, such as VMware CloudHealth, Apptio Cloudability, and CloudCheckr. Cost efficiencies can also be achieved with Compute Savings Plans, Reserved Instances, Spot Instances for ephemeral workloads, and Amazon CloudFront Security Savings Bundle. Continually reviewing cost metrics can help to identify over purchased or underutilized savings mechanisms. For example, you can optimize your storage costs with S3 Intelligent-Tiering, Amazon S3 Glacier, or implementing lifecycle policies and purge processes. Centralize redundant or shared infrastructure to optimize costs. Manage demand and supply resources dynamically by implementing scheduled or automatic scaling, buffering, or throttling. Review new EC2 instance types as they are released to take advantage of a better price-performance ratio.