Container Usage Report

Kubernetes Overview Report

Have you configured your reports? Report configuration helps you derive better insights quickly in the VMware Tanzu CloudHealth platform. See Configure Container Infrastructure for Cost Analysis.

It can take up to 24 hours after Collector deployment to start seeing meaningful reporting in the platform.

Kubernetes costs and usage visibility is critical for organizations while planning for chargeback/showback workflows or identifying cost or usage spikes in the environment. Tanzu CloudHealth brings together key reports in the platform to a central location with Kubernetes Overview. These usage reports are applicable for any Kubernetes environment, including GCP and Data Center. Kubernetes cost reporting that associates the Kubernetes usage data to the cloud bill is supported only for AWS, Azure and GCP.

To access Kubernetes Overview report on the platform, navigate to Reports > Containers > Kubernetes Overview.

kubernetes-overview-report

What Is the Kubernetes Overview Report

The Kubernetes Overview Report brings several key reports into a centralized dashboard enabling you to understand the cost and usage trends in aggregate, per cluster and namespace.

How to Use the Kubernetes Overview Report

At the top of the report, you can view the data segmented by cluster or by namespace.

Based on your selection, you will see an executive summary, cluster/namespace snapshot, containers cost history and container usage charts on the primary page.

From there, you can drill down into any cluster or namespace by selecting a cluster or namespace from the Snapshot table.

kubernetes-overview-drilldown

In this scenario, us1-prod cluster was selected, where we see a filtered executive summary, cluster details, and usage reports. In this view, you can better understand the usage trends over time for this cluster.

Upon clicking the View icon, the system will link out to the underlying OLAP chart, where you can leverage many more filters, categorizations, and continue to explore.

Glossary

Executive Summary Tiles

  • Clusters: The number of clusters Tanzu CloudHealth is collecting data from via the Tanzu CloudHealth collector for the day.
  • Namespaces: The day’s count of namespaces Tanzu CloudHealth is collecting via the Tanzu CloudHealth collector.
  • Nodes: The day’s count of nodes Tanzu CloudHealth is collecting via the Tanzu CloudHealth collector for the day.
  • Pods: The day’s count of pods Tanzu CloudHealth is collecting via the Tanzu CloudHealth collector.
  • Cost (MTD): The aggregate total cost of your Kubernetes clusters as per the configured Clusters Perspective in your Settings.

The count for Clusters, Namespaces, Nodes, and Pods are reset at the end of the day.

Cluster/Namespace Snapshot

  • Cluster Name: The user defined name for the cluster.
  • Cluster ID: The cluster ID for the cluster.
  • Clusters: The most recent count of clusters the namespace spans (collected via the Tanzu CloudHealth collector).
  • Namespaces: The most recent count of namespaces in the cluster (collected via the Tanzu CloudHealth collector).
  • Nodes: The most recent count of nodes in the cluster/namespace (collected via the Tanzu CloudHealth collector).
  • Pods: The most recent count of pods in the cluster/namespace (collected via the Tanzu CloudHealth collector).
  • Status: The health of the Tanzu CloudHealth collector running on the cluster. It is important that all clusters are configured and are in a healthy state to gain proper visibility of your Kubernetes footprint.
  • Kubernetes Version: The version of Kubernetes running on the cluster.

Containers Reports

  • Containers Cost History: A daily cost report categorized by the cluster/namespace perspective selected in the Kubernetes Configuration page.
  • Kubernetes Resources - CPU: A daily CPU usage report categorized by cluster/namespace.
  • Kubernetes Resources - Memory: A daily memory usage report categorized by cluster/namespace.

Cluster Resource Utilization Report

Analyze cluster usage to determine if you have over-provisioned resources in your cluster. For help reading and understanding the cluster resource report, see Understand Container Report Measures.

Have you configured your reports? Report configuration helps you derive better insights quickly in the Tanzu CloudHealth platform. See Configure Container Infrastructure for Cost Analysis.

It can take up to 24 hours after Collector deployment to start seeing meaningful reporting in the Tanzu CloudHealth platform.

Kubernetes dimensions with over 3,500 members are grouped into an Other (#) category, where (#) indicates the number of members being grouped in the particular view.

Cluster Resource Utilization

You can analyze Containers Resource utilization reports under:

  • Reports > Containers > Kubernetes Resources
  • Reports > Containers > ECS Resources

Insights This Report Provides

  • Is the cluster too big?
  • Are there excess provisioned resources in the cluster compared to those allocated to workloads?
  • Does the cluster resource mix match the workload resource needs?

Report Default Settings

  • Interval: Hourly
  • X-Axis: Hours
  • Y-Axis: (i) Available CPU Hours and (ii) Requested CPU Hours
  • Chart Type: Area

Examples

1. Kubernetes Cluster Resources over time (Default View)

k8 resource default

You can use this report to determine if you have over-provisioned resources in your cluster. Consider using this insight to start autoscaling the cluster.

2. CPU utilization by container tasks for Kubernetes Resources

Report Settings:

  • Interval: Daily, X-Axis: Days
  • Y-Axis: Cluster % CPU Used
  • Category: Perspective of Container Tasks
  • Chart Type: Bar - Stacked

Example 1 for K8 Resource CPU Utilization

Variation: Swap out Cluster % CPU Used on the Y-Axis with Relative Usage CPU to analyze which of your container tasks are consuming the most resources. While this report does not attach a monetary value to the utilization, it gives you an overview of how resources are being consumed across your infrastructure.

Example 2 for k8 resource CPU utlization

Container Resource Allocation Report

Analyze container resource allocation to understand which teams are using resources. For help reading and understanding the container resource allocation report, see Understand Container Report Measures.

Have you configured your reports? Report configuration helps you derive better insights quickly in the Tanzu CloudHealth platform. See Configure Container Infrastructure for Cost Analysis.

It can take up to 24 hours after Collector deployment to start seeing meaningful reporting in the Tanzu CloudHealth platform.

Container Resource Allocation

You can analyze reports relevant to Containers Resource allocation at the following tabs:

  • Reports > Containers > Kubernetes Allocation
  • Reports > Containers > ECS Allocation

Insights This Report Provides

  • Resource consumption to help match the business value
  • Relative resource consumption of different teams
  • Share of cluster resources consumed by different services that are owned by the same team

Report Default Settings

  • Interval: Daily
  • X-Axis: Days
  • Y-Axis: Requested CPU hours
  • Category: Cluster
  • Chart Type: Bar-Stacked

Examples

Cluster allocation by Days and Cluster for Kubernetes Allocation (Default View)

example-k8-allocation

Use this report to gain visibility into usage by team or service trended over time.

Understand Container Report Measures

What Are Report Measures

Measures in Container Reports allow you to analyze container data in meaningful ways. To access these measures, in any container report, expand the Y-Axis dropdown.

container report measures

Types of Measures

Tanzu CloudHealth reports on CPU and memory resource types for both Kubernetes and ECS.

Kubernetes Measures

  • Number of Nodes: The total number of nodes available in the cluster.
  • Available Resource: The average quantity of the resource for the reporting period. For example, the Available CPU for a server with 1 CPU available for 1 day would be 1 CPU for a day or 0.14 (1/7th) CPUs for a week.
  • Available Resource Hours: The resources available to the orchestrator. The time element accounts for the dynamic nature of resource availability. For example, the Available CPU Hours for a server with 1 CPU that was part of the cluster for 1 day is 24.
  • Cluster % Resource Requested: The resource allocated to a set of containers as a percentage of the overall cluster capacity for that resource over the reporting period. For example, on a given day in a cluster with 24 CPU hours, if a set of containers were requested and allocated 12 CPU hours, the Cluster % CPU for that container would be 50%.
  • Relative Usage Resource: The relative resource allocation among the different categories in the report. This representation can be helpful when comparing resource allocation to different container groups.
  • Requested Resource: The resources requested by the pods that define the amount of resource each container might need. For more information on Kubernetes Resource requests, refer to Managing Resources for Containers.
  • Resource Limits: The limits of the Kubernetes resources. For example, if a pod’s CPU limit is set to 1 CPU and the pod ran for one day, the daily report reports the CPU limit as 1 CPU for that specific day. To learn more about limits, see How Pods with Resource Limits are Run.
  • Resource Limit Hours: The amount of the resource used within the set limit, over the time period of the report. For example, if a pod’s CPU limit is set to 1 CPU and the pod ran for one day, the daily report reports the CPU limit hours as 24 CPU for that specific day.
  • Requested Resource Hours: The resource hours requested by the pods that define the amount of hours each container might need.
  • Resource Usage (Beta): The amount of the Kubernetes resource used.
  • Resource Utilization Percentage (Beta): The maximum amount of Kubernetes resources used divided by the resource request hours, displayed in percentage.

ECS Measures

  • Number of Nodes: The total number of nodes available in the cluster.
  • Available Resource: The average quantity of the resource for the reporting period. For example, the Available CPU for a server with 1 CPU available for 1 day would be 1 CPU for a day or 0.14 (1/7th) CPUs for a week.
  • Available Resource Hours: The resources available to the orchestrator. The time element accounts for the dynamic nature of resource availability. For example, the Available CPU Hours for a server with 1 CPU that was part of the cluster for 1 day is 24.
  • Cluster % Resource Registered: The resource allocated to a set of containers as a percentage of the overall cluster capacity for that resource over the reporting period. For example, on a given day in a cluster with 24 CPU hours, if a set of containers were requested and allocated 12 CPU hours, the Cluster % CPU for that container would be 50%.
  • Registered Resource: The resources requested by the pods that define the amount of resource each container might need.
  • Relative Usage Resource: The relative resource allocation among the different categories in the report. This representation can be helpful when comparing resource allocation to different container groups.
  • Relative Registered Resource: The relative resource allocation among the different categories in the report. This representation can be helpful when comparing resource allocation to different container groups.
check-circle-line exclamation-circle-line close-line
Scroll to top icon