A committed use discount (CUD) from Google requires you to make a commitment to a minimum level of resources over a one- or three-year period. In exchange, you’ll receive a discounted price—with savings up to 57 percent for most resources, and possibly even more for virtual machines.
You’re billed a monthly fee, based on the product pricing at the time of purchase. You pay the same monthly fee for the duration of your commitment period, even if pricing on that product goes up. Once you purchase a CUD, however, you can’t cancel it.
Unlike similar discount offerings from other cloud providers, you can’t pay the entire amount upfront. And, if your usage exceeds the amount you’ve committed to in the CUD, the overage will be charged at the standard rate for that product. Keep in mind that you also can’t roll over unused commitments.
So when should you consider a committed use discount?
Making a commitment
There are two main types of committed use discounts: spend-based or resource-based. Here’s how it works:
Spend-based commitments: This requires you to commit to product pricing over a one- or three-year period. Your spend is measured in dollars per hour; anything over your committed amount is charged at the standard rate. You can buy spend-based CUDs for any project in your Cloud Billing account.
Spend-based CUDs are available for: AlloyDB for PostgreSQL, Cloud Bigtable, Cloud Run, Cloud Spanner, Cloud SQL, Compute Engine, Google Cloud VMware Engine, Google Kubernetes Engine, and Memorystore. Google provides a CUD analysis dashboard to help you evaluate the financial impact of your committed use discounts.
Resource-based commitments: In this case, a discount is provided in exchange for using a minimum level of Compute Engine resources in a particular region. Unlike spend-based commitments, resource-based commitments are typically purchased for a specific project, and they’re purchased in terms of the underlying vCPU, GPU, and local SSD resources.
Consider this if you have predictable Compute Engine needs, but still want flexibility in how you use machine instances. Keep in mind that committed use discounts apply at the project level by default, where you enrolled in them; however, you can share committed use discounts across projects that are under the same Cloud Billing account by enabling CUD sharing.
A CUD for every cloud service
Different products and services have different CUDs. Here’s how they’re applied to specific cloud services:
AlloyDB for PostgreSQL: Applies to all AlloyDB for PostgreSQL instance vCPU and memory usage in any project or region in a single Cloud Billing account, but not storage, backup, or network egress.
Cloud Spanner: Applies to all Cloud Spanner compute capacity associated with a single Cloud Billing account, whether configured as single-region or multi-region instances, but not storage, backup or network egress).
Cloud SQL: Applies to all Cloud SQL database instance vCPU and memory usage in the region you purchased the commitments; this includes MySQL, PostgreSQL, and SQL Server, but not shared CPU machine types, persistent disk snapshots, storage, IP addresses, network egress or licensing.
Google Kubernetes Engine: Applies to all Autopilot Pod workload vCPU, memory, and ephemeral storage usage in the region you purchased commitments, but not to the cluster management fee or GKE Standard mode compute nodes.
Google Cloud VMware Engine: Applies to aggregate VMware Engine node CPU and memory usage in the regions where the service is available.
Getting the most out of CUDs
Maximizing your cost savings from committed use discounts often requires upfront work. For example, since different services have different CUDs, optimizing these discounts takes some planning and strategizing. Oftentimes, the best approach is to first obtain the resources, purchase a CUD and then use the Google Cloud reservation feature to reserve it.
To determine which workloads are potential candidates for CUDs, you’ll need to analyze usage trends and metrics in your cloud management platform. It’s also possible to use Google’s committed use discounts with other pricing models, so you may want to consider CUDs as part of your larger FinOps strategy.
If you’re looking to take advantage of CUDs, or better manage costs in the cloud, our team can help. As a premier Google Cloud managed service provider, our team can provide end-to-end services for all Google Cloud solutions. So whether you’re looking for support with CUDs, FinOps, or are looking for dedicated consulting support, our team can help.
Get in touch Pythian to learn more about committed use discounts and how we can help.