GCP Rightsizing

Benefits of GCP Rightsizing

The two key benefits of rightsizing are infrastructure optimization and cost reduction. During a rightsizing analysis of their infrastructure, organizations discover assets that can be downsized or terminated to save money or upgraded to improve performance.

Downsizing

An asset is underutilized and a strong candidate for downsizing if it exhibits low utilization (less than 20%, for example) for core performance metrics. In this case, the best practice is to downgrade the asset to a smaller footprint.

Terminating

Your cloud infrastructure will likely contain assets that are running but not being used. These assets are called zombies, and they are good candidates for termination. Zombies result when someone forgets to turn the assets off after use or when the asset fails due to script errors. Regardless of the cause, cloud providers continue to charge for these unused assets because they are in a running state. You can reduce costs by proactively identifying and terminating these assets.

Upgrading

By downgrading underutilized assets and terminating unused ones, you can optimize for performance as well as reduce costs. Upgrading assets, on the other hand, results in increased spend. However, by upgrading you can ensure that your assets are able to meet surges in demand.

GCE Rightsizing Reports

Insights These Reports Provide

The GCE Rightsizing reports provide the following insights:

• How well all your GCP Projects’ compute instances are being utilized in terms of the workloads you are running on them.
• Opportunities for rightsizing all GCP projects’ compute instances, thereby saving costs and optimizing workloads.
• Visualization of compute instance utilization across all Google projects.

Source of Rightsizing Recommendations

VMware Tanzu CloudHealth gathers Google rightsizing recommendations through API calls. For more information on Google rightsizing recommendations, see Applying Sizing Recommendations for Google VM Instances.

Compute instance metrics are retrieved for the previous 8 days.

Source compute instances must be active for at least 24 hours to begin generating rightsizing recommendations. Tanzu CloudHealth recommends waiting at least 8 days to generate rightsizing recommendations on new compute instances to ensure more accurate recommendations.

How to Interpret the GCE Rightsizing Report

This report highlights underutilized and overutilized GCP compute instances. To access the GCE Rightsizing report, go to Recommendations > Rightsizing > GCE Rightsizing.

GCE Recommendation Summary

The GCE Rightsizing Recommendation Summary displays your total number of opportunities to rightsize, as well as your number of opportunities to improve performance and to save on costs. The recommendation summary also displays your projected total monthly savings if you implement all the recommendations in the report.

Hover the mouse over the How did we calculate this link to view a breakdown of savings based on different rightsizing recommendations.

This calculation does not include the GCP Sustained User Discount (SUD).

Recommendation Details

To view a drilldown into a recommendation, select that recommendation from the GCE Rightsizing report. The Recommendation Details dialog box appears. The Recommendation Details dialog box contains additional details about the recommendation, such as the date and time the recommendation was originally generated.

To implement the rightsizing recommendation, copy the text in the gcloud CLI Command field. In the Google Console, stop the instance and then paste copied text in the gcloud CLI. The instance is then upgraded or downgraded to the recommended machine type.

How to Interpret Age (Days)

Age (Days) is the number of consecutive days that Tanzu CloudHealth has collected the Google rightsizing recommendation for the current machine type. Rightsizing recommendations that have been recommended for multiple days are more ‘stable’ and less likely to be based on a fluke in metrics utilization.

Tanzu CloudHealth can only collect Google rightsizing recommendations after you configure your GCP account with Tanzu CloudHealth. Therefore, it’s possible that Google has made the same recommendation for longer than the Age (Days) column states. For example, if Google has made the same rightsizing recommendation for 30 days but you only configured Tanzu CloudHealth with your GCP account 15 days ago, the Age (Days) column dates the recommendation as 15 days old.

How to Interpret Savings and Additional Cost

• Savings are the estimated monthly savings that you gain by implementing the rightsizing recommendations for an underutilized GCP compute instance by downgrading to a less expensive machine type.
• Additional Cost is the estimated monthly additional cost you gain by upgrading an overutilized compute instance to a more expensive machine type that can better handle the instance’s metrics requirements.

How to Interpret Recommendations

Recommendations is the course of action that Google rightsizing recommendations calculates after analyzing GCP project utilization. Depending on whether the compute instance is currently under- or overutilized, this recommendation may be to downgrade or upgrade. For more information, see Applying Sizing Recommendations for Google VM Instances.

How to Interpret Justification Summary

Justification Summary is the reason for upgrading or downgrading your instance’s machine type according to the rightsizing recommendation.

Kubernetes Rightsizing for GCP

The Kubernetes Rightsizing tool can help you use your Kubernetes Containers more efficiently and save money. Using Rightsizing, you can:

• Analyze the workload patterns of your containers
• Apply custom efficiency targets to generate recommendations for optimal container sizes and to meet different application and enterprise requirements

Unlike the legacy Rightsizing tool, you can change efficiency targets and instantly view updated recommendations to ensure the best fit.

To access the Kubernetes Rightsizing tool, navigate to Recommendations > Rightsizing > GKE Rightsizing.

Using the Rightsizing Dashboard

The Kubernetes Rightsizing dashboard provides information about the efficiency of your containers, displays the status of each and provides recommendations.

The Efficiency Summary provides the following information:

• Current CPU: The total requested compute capacity across all your containers (in CPU cores)
• Recommended CPU: The total recommended compute capacity across all your containers (in CPU cores)
• Current Memory: The total requested memory capacity across all your containers (in GB)
• Recommended Memory: The total recommended memory capacity across all your containers (in GB)
• No Change: Total number of containers that do not need any change and thus, have no recommendations.
• Resize: Total number of containers that need to resize their CPU and/or Memory requests and thus, have recommendations.

You can set an efficiency target and date range to help generate targeted recommendations. For more information about creating and using efficiency targets, see Efficiency Targets.

Click Filters to view more filter options for your recommendations, such as clusters, namespaces and workload types.

Click Export to download a CSV file of the recommendations.

Understand Recommendations

The Kubernetes Recommendations table shows all containers that fit the selected filters and that have a recommendation. You can also enable Show all recommendations to view containers that do not currently have a recommendation.

The dashboard displays the number of containers with recommendations and their fit:

• Good fit: The container meets all efficiency targets.
• Under target: The container is underutilized and does not meet all efficiency targets. Tanzu CloudHealth provides a recommendation to optimize the container.
• Over target: The container is overutilized and exceeds one or more efficiency targets.
• Undetermined fit: One or more of the metrics is above target, but one or more of another metric is below target, so the fit could not be determined.
• Not enough data: There was no data available for a metric for the selected time period, so the fit could not be determined.

The Recommendations table also displays the following information:

• CPU Requests: The requested compute capacity for a container.
• Memory Requests: The requested memory capacity for a container.
• Recommended CPU Requests: The recommended compute capacity for a container based on applied efficiency target.
• Recommended Memory Requests: The recommended memory capacity for a container based on applied efficiency target.

Resize Recommendations

Identifies the right amount of compute and/or memory requests that can handle the usage for the current container.

If Tanzu CloudHealth is unable to make a recommendation for a container, it is hidden from the dashboard. You can enable Show all recommendations to view containers that do not currently have a recommendation.

Note: No recommendations are generated in the following situations

• If the container has no options matching the efficiency target requirements
• If metric data is unavailable for all metrics in the selected efficiency target

View Recommendations

To view the recommendation options and metrics for a specific container, click the container name on the Rightsizing page. On the recommendation details page, you can see current container’s compute and memory requests, and the recommended requests for both.

The Resource Metrics charts show current performance metrics and the target efficiency range. You can use these charts to determine how your current usage compares to the target metrics.

The Kubernetes Rightsizing tool also gives you an option to view recommendations grouped by -

• Namespace
• Cluster
• Perspective
• Workload

Example - From the Group By dropdown, select Workload and click Update. You will see all the recommendations grouped by Workload. Expand a workload name to view all the containers that fall into that specific account. Click the container name to know the recommendation details of the selected container.

Efficiency Targets

Efficiency targets are a collection of a azure resource usage attributes and utilization ranges that can be applied to meet application or enterprise requirements. Tanzu CloudHealth’s new rightsizing report supports the following GCP resources -

• Containers

Rightsizing includes two default efficiency targets: Average Metrics and Maximum Metrics. You can also define your own targets by creating a custom efficiency target.

Create a Custom Efficiency Target

To create a custom efficiency target:

1. From the Actions menu, select Add custom efficiency target.
2. Provide a name for your new efficiency target.
3. Select one or more metrics to set as a target.
4. Select whether the metric should be defined by average performance (AVG), or limit the metric from going beyond a maximum value (MAX).
5. Define the target threshold range in values from 1 to 100.

View, Edit or Delete a Custom Efficiency Target

To view, edit, or delete an existing custom efficiency target:

1. From the Actions menu, select Manage custom efficiency targets.
2. Select the custom efficiency target from the Efficiency target dropdown.
3. Make changes to the custom efficiency target:
   • To edit the custom efficiency target, modify any metrics and select Save.
   • To delete the custom efficiency target, select Delete.