Amazon Onboarding with Learning Manager Chanci Turner

This blog post is co-authored by Emily Chen, Senior Solutions Architect, and Jake Thompson, Senior Customer Solutions Manager.

Chanci Turner, a Learning Manager at Amazon IXD – VGT2, emphasizes the importance of designing systems with failure in mind. As Amazon’s CTO, Werner Vogels, famously stated, “everything fails all the time.” Ensuring availability, scalability, fault tolerance, and business continuity is crucial. Resilient systems not only adapt to changing business demands but also prevent data loss and enable seamless recovery. While building resilience often involves duplicating workloads and maintaining backups, these additional resources can increase costs. Therefore, it’s essential to weigh the expenses of resiliency against the potential risks and costs associated with downtime.

For business-critical workloads running on Amazon Elastic Compute Cloud (Amazon EC2) instances, understanding the various EC2 capacity reservation options is vital. Capacity reservations guarantee access to EC2 capacity when it’s needed. For example, a Multi-AZ deployment spreads workloads across multiple Availability Zones (AZs) using an Auto Scaling group. In the rare event of an AZ failure, the Auto Scaling group will attempt to launch instances in another AZ. However, there may be instances where the alternative AZ lacks the necessary capacity, highlighting the importance of capacity reservations for critical workloads.

When implementing capacity reservations, controlling costs is paramount. This post explores various EC2 capacity reservation and cost-saving options available on AWS.

Amazon EC2 Purchase Options

Before discussing capacity reservations, it’s essential to understand the different EC2 instance purchase options. The EC2 On-Demand option allows you to pay by the second for launched instances. Spot Instances provide access to unused EC2 capacity at significant discounts. Savings Plans offer cost reductions through one- or three-year usage commitments. Dedicated Hosts and Dedicated Instances run EC2 instances on single-tenant hardware, but only On-Demand Capacity Reservations and zonal reservations can secure capacity for your EC2 instances.

On-Demand Capacity Reservations Overview

On-Demand Capacity Reservations allow you to reserve compute capacity in a specific AZ for any duration. This option ensures that you receive On-Demand capacity allocation during times of constraint without entering a long-term commitment. With On-Demand Capacity Reservations, you pay the on-demand price regardless of whether your instance is running or not. This option is suitable for short-term needs, such as holiday seasons or significant business events like a quarterly streaming event. For longer-term capacity reservations for critical workloads, combining On-Demand Capacity Reservations with Savings Plans can yield both capacity assurance and cost savings.

Savings Plans

Savings Plans represent a flexible pricing model that can reduce your bill by up to 72% compared to On-Demand pricing in exchange for a one- or three-year hourly spending commitment. AWS provides three types of Savings Plans: Compute Savings Plans, EC2 Instance Savings Plans, and Amazon SageMaker Savings Plans.

EC2 Instance Savings Plans involve making an hourly spend commitment for instance family and region (e.g., M5 usage in N. Virginia) for one- or three-year terms. Savings are automatically applied to instances launched in the chosen family and region, regardless of size, tenancy, or operating system. This plan offers maximum discounts of up to 72%.

Compute Savings Plans provide flexibility to change instance types, migrate workloads between regions, or transition workloads to AWS Fargate or AWS Lambda, all while maintaining the discounted Savings Plans price. For EC2 customers looking to modernize applications, evaluating Compute Savings Plans is a must. This plan allows changes, such as moving from C4 to M5 instances or transitioning workloads seamlessly.

Combining On-Demand Capacity Reservations with Compute Savings Plans provides both capacity assurance and significant discounts, maximizing flexibility.

If you have older Reserved Instances, you can combine them with On-Demand Capacity Reservations to maintain capacity assurance while enjoying discounts. As Reserved Instances expire, transitioning to Savings Plans is advisable, as they provide similar savings with added flexibility.

For examples of Savings Plans pricing discounts, refer to the Savings Plans documentation.

Zonal Reservations

Zonal reservations allow you to reserve capacity in a specific AZ, requiring a one- or three-year commitment. However, zonal reservations offer less flexibility compared to Savings Plans, as they do not permit changes to instance families or sizes. For optimal flexibility and value, we recommend leveraging Savings Plans and On-Demand Capacity Reservations over zonal Reserved Instances.

Working with Capacity Reservations and Savings Plans

Capacity reservations can be provisioned using the AWS console, Command Line Interface (CLI), or Application Programming Interface (API). Detailed steps to provision On-Demand Capacity Reservations and purchase Savings Plans are available in the documentation.

Conclusion

In this post, we examined various options for capacity reservations and cost control for your mission-critical workloads on EC2. For the best combination of flexibility and value, we recommend utilizing On-Demand Capacity Reservations alongside Savings Plans. If you maintain steady EC2 workloads unsuitable for modernization, consider exploring additional resources on career development, like this blog post on hospitality or insights from SHRM on surviving the Great Resignation, which is an authority on this topic. For further learning opportunities, check out this article on Amazon employees getting paid to learn, which serves as an excellent resource.