One of the greatest advantages of the Cloud is how easy it can be to scale your usage up or down. However, if you're working with a predictable compute workload, you can save some money with committed use discounts. In this video, we'll talk about what committed use discounts are and how you can use them to optimize your costs. Committed use discounts or CUDs for short, are basically contracts saying you'll commit to a certain amount of Compute Engine resources for one or a three-year period. In exchange for your commitment, you'll get a significant price discount even as high as 70 percent in some cases. You can make these commitments on your typical compute resources such as vCPUs and memory, but also on GPUs and local SSDs. It's worth noting that committed use discounts are different from sustained use discounts. Committed use discounts can be purchased before consuming Compute Engine resources based on what you intend to use. Sustained use discounts are automatic discounts credited to your account after running specific Compute Engine resources for a significant portion of the billing month. Remember, the two discounts are different and they can't be combined. Let's look at how to set up commitments. Then we'll explore the committed use discount analysis report to help you gain some insights from your data. The committed use discount analysis report is meant to help you answer questions like, how much of my eligible Compute Engine usage is covered by my commitments? How much are my commitments actually saving me? I'm I fully utilizing my existing commitments? To purchase commitments refer to the Google Cloud Console. Using the top-left navigation, click on "Compute Engine" and select "Committed Use Discounts". From here, click on "Purchase Commitment" to enter the specifics, such as the timeframe and which resources you want to purchase. Each commitment is purchased for a specific project and region. Commitments for vCPUs and memory are applied simultaneously, but you have separate commitments for GPUs and local SSDs as well. You'll also get an estimated cost to let you see how much you'll pay each month and how much the commitment would save you. If you run into any issues purchasing a commitment, you may need to make sure you have the right quota for it. Once you've made the commitment, you'll start to pay for that commitment over its lifetime with each billing cycle, even if you don't have any resources active. In this example, a certain amount of vCPUs and RAM had been used by Compute Engine instance. If you have a lot of resources and commitments, it can be confusing to see what you're spending and which commitments are active, especially since they don't auto renew. That's where the committed use discounts analysis report comes in. It helps you visualize all of your commitments alongside your compute resources. Using the Google Cloud console, choose a billing account and click on the commitments option to get the report. Unlike billing reports, you won't be able to get access to the CUD analysis report by being a project owner so you can filter by project. You'll need to be a billing account administrator or billing account viewer to see this report. Let's step through the different areas to see how this report works. The resources are split up by type. This lets you see vCPU, RAM, SSD, and each GPU types separately, along with your usage and relevant commitments. You can toggle between them by clicking the tabs here. We'll start with vCPUs. Above the graph, you can see a summary of your commitments in one row. The left side, let's you see if you're looking at all regions or if you're looking at specific regions, which I'll show you how to do in just a second. The middle shows the current number of active commitments, which may take a day or two to update when a new commitment is purchased. The right side shows you the utilization of your commitments by comparing the amount of eligible usage being covered by your commitments to the total commitments you have. You can also hover over the question mark for the specific resource counts. The stacked bar chart itself shows you the resource usage for each day, the amount that was covered by a commitment, and a line that shows you the current quantity of commitments. You can hover over each column to see the exact usage numbers for each day. There's also a summary table below the chart. When looking at the CUD report for an entire billing account, the CUD utilization is grouped by projects and regions. That's why for certain time-frames, you might notice a difference between eligible resource usage compared to CUD coverage. To understand the Delta, you'll need to break down the data by region and then project to see why there was a dip like that. Let's look at a specific subset of days as an example. Here, the usage view is narrowed down to a few days. On the first day, all commitments were fully utilized and there was additional eligible resource usage on top of the purchase commitments. For the remaining days, the graph is revealing different information. The dotted line that shows the active commitment hasn't changed, but the actual resource usage has. Since we're talking about the Cloud, it makes sense for the resource usage to fluctuate, but how can we find out why the resource usage wasn't covered by the commitments? Let's drill down using the filters. You can also change the time range here, if you want to see more dates. We'll change the view to region and see the specific split up by each region. This is the US Central 1 region, but the graph is still showing the same usage pattern. Remember that committed use discounts are specific to a project and region. Let's use the filters to select a specific project. With this filtered view, we're now able to answer our question. Fewer vCPUs were used for a few days in this project which led to commitment under-use. Understanding more details about your commitment utilization can give you a much better idea of how to plan for future usage of your resources and how to ensure purchase commitments are fully utilized, or even where you may need to purchase additional commitments. Committed use discounts can help you save money on your Compute Engine resources, especially if you have a lot of consistent usage. It's also important to use the committed use discounts analysis report to see how your existing commitments are sizing up and if it makes sense to purchase additional commitments. Use the GCP console and try the report for yourself.
