How to get more value from your cloud commitments

If you are familiar with the cloud, you already know that the big hyperscalers offer a set of pricing models that provide varying levels of flexibility and discounts. The most common models are On-Demand (OD), Commitments—such as Reserved Instances (RI), Savings Plans (SP), Committed Use Discounts (CUD), and preemptive instances like spot instances/spot VMs.

Commitments appeal to organizations as they provide a significant discount in exchange for committing to a specific amount of compute capacity for a set term. While commitment plans can significantly reduce costs, utilizing them effectively in a dynamic compute environment can be challenging.

Organizations need to ensure that commitments are not left underutilized or not being used optimally across their account while resources such as on-demand or spot instances are purchased to cover momentary needs instead. Such underutilization and sub-optimal coverage of commitments incurs unnecessary costs and limits any savings achieved.

In this blog, we will provide actionable steps to get started quickly, demonstrating how Spot by NetApp can automate and optimize your commitment utilization as part of your infrastructure optimization processes. Ready to harness the true power of your cloud commitments? Let’s dive in.

Optimal commitments usage and coverage

To ensure optimal commitment utilization for VM or containerized workloads, organizations should constantly perform real-time inventory assessments of the organization’s commitment assets and the instances that are running, to identify opportunities for replacing spot or OD instances with underutilized committed capacity. Since resource utilization is dynamic, continuously monitoring and adjusting commitment utilization and assigning them is hard to do, which leads to wasted resources and increased costs.

Spot Ocean and Elastigroup automatically optimize commitments utilization inherently as part of their infrastructure optimization processes by ensuring that purchased commitments are prioritized, and the greatest cost savings possible are achieved by dynamically and efficiently allocating resources.

This is achieved by:

• Prioritizing commitments usage: Existing commitments are mapped by Ocean and Elastigroup and prioritized for consumption over on-demand or preemptive instances, to avoid additional cost.
• Increase commitments coverage: Whenever an instance is using a reservation which is needed elsewhere in the account, Ocean and Elastigroup will seamlessly switch to spot instances, freeing up this reservation to be used where it is needed most. In addition, Ocean and Elastigroup continuously monitor the entire deployment to identify underutilized commitments so that can now be used to replace any on-demand instances. These actions reduce the amount of on-demand compute used and prevent instances from being locked into commitments when a cluster is set to run on spot instances. This way, workloads requiring on-demand instances anywhere in the account can still use reservations, optimizing resource allocation and cost efficiency.
• Granular reservation usage control at the Virtual Node Group (VNG) level: For users working on Kubernetes, Ocean provides granular control by enabling or disabling reservation usage at the VNG level. This empowers organizations to tailor reservation allocation according to their specific workload requirements, providing unparalleled flexibility and resource management.

Maximize savings with commitment optimization

The savings achieved through commitment optimization varies, but let’s explore an example to illustrate the impact of this capability. Imagine an organization with multiple clusters. One of its clusters (let’s call it #1) is set to run 100% on spot instances, but instead it utilizes underutilized commitments. Another cluster (let’s call it #2) is set to run on-demand.

Monitoring the entire account, to maximize savings Spot Ocean will release the commitments from cluster #1 and move them to cluster #2 instead of the on-demand instances. Ocean will then assign spot instances to cluster #1. When the commitments are released from cluster #2, Ocean will again use them for cluster #1 and will stop using the spot instances. By dynamically utilizing underutilized commitments, Ocean reduces the reliance on full price on-demand and spot instances, thereby reducing costs and maximizing commitment utilization.

Spot Ocean dynamic commitments in action

Start maximizing your commitments utilization

Ocean and Elastigroup help organizations utilize the least amount of on-demand resources possible, ensuring optimal cost efficiency and resource allocation for VM, containers, and Kubernetes infrastructure. Ready to unlock the full potential of your commitments and optimize your resource allocation? Schedule your personalized demo today.