|
|
Today, I’m excited to announce Amazon EC2 Capacity Manager, a centralized solution for monitoring, analyzing, and managing capacity usage across all AWS accounts and regions from a single interface. This service aggregates capacity information with hourly refresh rates and provides optimization capabilities based on priority, streamlining capacity management workflows that previously required custom automation or manual data collection from multiple AWS services.
Organizations using Amazon Elastic Compute Cloud (Amazon EC2) at scale run hundreds of instance types across multiple Availability Zones and accounts using On-Demand Instances, Instant Instances, and Capacity Reservations. This complexity means that customers currently access capacity data through a variety of AWS services, including the AWS Management Console, cost and usage reports, Amazon CloudWatch, and EC2. describe Apis. This distributed approach can create operational overhead through manual data collection, context switching between tools, and the need for custom automation to aggregate information for capacity optimization analysis.
EC2 Capacity Manager helps you overcome these operational complexities by consolidating all capacity data into a single dashboard. You can now view cross-account and cross-region capacity metrics for on-demand instances, instant instances, and capacity reservations across all AWS commercial regions from one place, eliminating the need to build your own data collection tools or move between multiple AWS services.
This consolidated visibility can help you discover cost savings by highlighting underutilized capacity reservations, analyzing usage patterns across instance types, and providing insight into Spot Instance outage patterns. With access to comprehensive capacity data in one place, you can make more informed decisions about setting up your infrastructure right and optimizing your EC2 spend.
Let me show you the capabilities of EC2 Capacity Manager in detail.
Getting started with EC2 Capacity Manager
In the AWS Management Console, I go to Amazon EC2 and select Capacity Manager from the navigation panel. I enable EC2 Capacity Manager through service settings. The service aggregates historical data from the previous 14 days during initial setup.
Hand Dashboard displays capacity utilization across all instance types through a comprehensive overview that presents key metrics at a glance. Capacity overview cards for Reservation, Useand Place view trend indicators and percentage changes to help you quickly identify capacity patterns. You can apply filtering using the date filter controls, which include selecting a period, configuring a time zone, and setting an interval.
You can select different units to analyze data by vCPU, number of instances, or estimated cost to understand resource consumption patterns. Estimated costs are based on published On-Demand rates and do not include savings plans or other discounts. This pricing benchmark helps you compare the relative impact of unused capacity across different instance types – for example, 100 vCPU hours of unused p5 reservations has a greater cost impact than 100 hours of unused t3 vCPU reservations.
The dashboard includes a detailed Usage metrics both with visualization of total usage and with graphs of usage over time. The total usage section shows the breakdown between reserved usage, unreserved usage and spot usage. The Usage Over Time graph provides a visualization that tracks capacity trends over time, helping you identify usage patterns and peak periods.
Below booking metrics, Reserved capacity trends visualizes used and unused reserved capacity in a selected period and shows the proportion of reserved vCPU hours that remain unused compared to those actively consumed, helping you track patterns in reservation efficiency and identify periods of consistently low utilization. This visibility can help you reduce costs by identifying unused bookings and helping you make informed decisions about capacity adjustments.
Tea Unused capacity section lists unused capacity reservations by instance type and availability zone combinations and shows specific utilization percentages and instance types in different availability zones. This prioritized list will help you identify potential savings with a direct overview of the cost of unused capacity.
Tea Use tab provides detailed historical trends and usage statistics across all AWS regions for Instant Instances, On-Demand Instances, Capacity Reservations, Reserved Instances, and Savings Plans. Use of dedicated hosts is not included. Tea Dimensional filter helps you group and filter capacity data by account ID, region, instance family, availability zone, and instance type, and create custom views that reveal usage patterns across your AWS accounts and organizations. This helps you analyze specific configurations and compare performance between accounts or regions.
Tea Aggregation section provides a comprehensive usage table across EC2 and Spot instances. You can select different units to analyze data by vCPU, number of instances, or estimated cost to understand resource consumption patterns. The table shows a breakdown of instance families with total usage statistics, reserved usage hours, unreserved usage hours, and spot usage data. Each row contains a View schedule action for detailed analysis.
Tea Capacity utilization or estimated cost trends section shows usage trends, reserved usage, unreserved usage and spot usage. You can filter the displayed data and adjust the unit of measurement to show historical patterns. These filtering and analysis tools help you identify utilization trends, compare costs across dimensions, and make informed decisions for capacity planning and optimization.
When you choose View schedule from Aggregation table, you will get to the details Distribution of use based on the dimension filters you selected. This breakdown shows usage patterns for each instance type within selected family and availability zone combinations to help you identify specific optimization opportunities.
Tea Reservation card displays capacity reservation usage with automatic analytics features that generate prioritized lists of opportunities to optimize. Similar to Use you can apply dimension filters by Account ID, Region, Instance Group, Availability Zone, and Instance Type, along with other options related to reservation details. You can view data for individual line items on each of the tabs. Specifically for reservations, you can view specific reservations and access detailed information about on-demand capacity reservations (ODCRs), including usage history, configuration parameters, and current status. If ODCR exists in the same account as Capacity Manager, you can edit reservation parameters directly from this interface, eliminating the need to navigate to separate sections of the EC2 console to manage reservations.
Tea Statistics section provides summary metrics, including total reservations, total utilization percentage, reserved capacity totals, used and unused capacity volumes, average planned reservations, and account counts, instance families, and reserved areas.
This consolidated view helps you understand the distribution of bookings and usage patterns across your infrastructure. For example, you may find that your development accounts consistently show 30% reservation utilization, while production accounts exceed 95%, indicating an opportunity to redistribute or adjust reservations. Similarly, you may find that particular instance families in certain regions have consistently low utilization rates, suggesting candidates for reservation adjustments or workload optimization. These insights help you make data-driven decisions about purchases, modifications or cancellations to better match your reserved capacity with actual usage patterns.
Tea Place focuses on Spot instance usage and shows how long your Spot instances are running before they are interrupted. This analysis of Spot Instance usage patterns will help you identify opportunities to optimize your Spot Instance workloads. To improve workload flexibility, you can use Spot location score recommendations.
For organizations requiring data export capabilities, Capacity Manager includes data export to Amazon Simple Storage Service (Amazon S3) buckets for capacity analysis. You can view and manage data exports via Data export a tab that helps you create new exports, track delivery status, and configure export schedules to analyze capacity data outside of the AWS Management Console.
Data exports extend your analytical capabilities by storing capacity data beyond the 90-day retention period available through the console and API. This extended retention enables long-term trend analysis and historical capacity planning. You can also integrate exported data with existing analytical workflows, business intelligence tools, or custom reporting systems to incorporate EC2 capacity metrics into broader infrastructure analysis and decision-making.
Tea Settings section provides configuration options for integrating AWS organizations, enabling centralized capacity management across multiple accounts. Organization administrators can enable enterprise-wide capacity visibility or delegate access to specific accounts while maintaining appropriate permissions and access controls.
Now available
EC2 Capacity Manager eliminates the operational overhead of collecting and analyzing capacity data from multiple sources. The service provides opportunities for automatic optimization, centralized visibility for multiple accounts and direct access to capacity management tools. You can reduce manual analysis time while improving capacity utilization and cost optimization across your EC2 infrastructure.
Amazon EC2 Capacity Manager is available at no additional cost. To get started using Amazon EC2 Capacity Manager, visit the Amazon EC2 console or access the service’s API. The service is available in all commercial AWS regions.
To learn more, visit the EC2 Capacity Manager documentation.
– Esra