What’s new in cost management

With a new cost model metric, you can now directly assign costs to your projects by using tag-based rates. After you apply a tag rate to a project, the monthly cost is assigned to that project. If the project has nodes associated with it, the cost is then distributed across those nodes. With this update, you can have better cost visibility, more precise financial tracking, and the ability to assign quota-based costs or minimum costs.

To get more precise control over your billing, you can now bill Virtual Machines (VMs) based on the number of physical VM cores that are utilized. This new VM core metric enables you to bill either hourly or monthly. Your system must be running on Operator Release 4.0.0 or later.

You can use a new cost model metric, cluster-hour`, to better model the hourly cost of operating your entire cluster. This metric adds additional hourly costs of each of your clusters to your overall cost. You can distribute costs based on either actual CPU, or memory usage, which ensures that workloads are proportionately charged based on the resources that they consume.

To view all metrics, see Cost model metrics.

Oracle Cloud is no longer a supported cloud provider for cost management integrations.

There are hidden costs associated with running an OpenShift cluster. Many of these costs are from subscriptions related to CPU core metrics.

To better capture these costs and distribute them to user projects, cost models have new metrics. To view the metrics, see Cost model metrics.

Cost management made several enhancements that are related to VMs:

When multiple VMs share storage, the storage costs are attributed to only one of the VMs. A planned enhancement will distribute the cost across all the attached VMs.

If you manually configured the cost management metrics operator to use basic authentication, you need to set up service account authentication.

For more information, see Configuring service account authentication for the cost management metrics operator and Transition of Red Hat Hybrid Cloud Console APIs from basic authentication to token-based authentication via service accounts.

With support for tracking Red Hat OpenShift Virtualization, you can now view costs that are directly associated with virtual machines (VMs) running on your OpenShift clusters. There is also a new API endpoint: /api/cost-management/v1/reports/openshift/resources/virtual-machines/.

To find your Red Hat OpenShift Virtualization data, click the Virtualization tab on the OpenShift details page.

To learn about Red Hat OpenShift Virtualization, see Virtualization.

Unattributed Storage is a type of project that gets created when cost management is unable to correlate a portion of the cloud cost to an OpenShift namespace.

This storage type was originally only available for Azure, but now it is also available for AWS. For more information, see Unattributed Storage project for Azure and AWS.

Based on customer feedback, the instructions for how to create an integration with Microsoft Azure were updated. The new version of the documentation reflects the latest changes to the Azure UI and aims to resolve customer pain points by being more clear and easy to follow.

There are two options when you create an integration: a basic path where you use unfiltered data, and an advanced path with filtered data. The instructions for both sections are updated.

To learn more about Azure and explore the new updates, go to Integrating Microsoft Azure data into cost management.

Unattributed Storage is a type of project that gets created when cost management is unable to correlate a portion of the cloud cost to an OpenShift namespace.

There are two scenarios where Unattributed Storage can happen:

To learn more about how to calculate Unattributed Storage in Azure, see Unattributed Storage project.

Costs associated with ingress and egress network traffic for individual nodes are now separated. A new project called network unattributed shows costs related to network traffic. You can use a cost model to distribute theses network costs in a similar way to platform and worker unallocated costs.

If you have an AWS savings plan for the EC2 instances running on OpenShift nodes, cost management uses the savings plan cost by default.

Customers noticed that cloud and OpenShift costs in cost management do not always match the data that is in the cloud provider’s console. Most of the time, this discrepancy is because there is a missing integration between the Hybrid Cloud Console and either the OpenShift cluster, or the cloud customer account. This missing integration creates discrepancies in billing and cluster data.

To help solve this discrepancy, we added two new views in the OpenShift details page of cost management:

You can now unify your tags by mapping ones that have similar or related meanings. This functionality offers greater flexibility and control over how you categorize and track your resources. Tag mapping reduces complexity, helps you have accurate reporting and analysis, and enhances consistency across your resources.

For example, imagine that you are tracking your environment spending across different platforms, each with its own naming conventions:

Platforms:

Variations in your key values:

With the tag mapping feature, you can seamlessly convert these key values under a single expected key, such as environment, to ensure that you are consistent in tracking your spending.

To learn about tag mapping, see Enabling and creating a tag mapping.

The Cost Management Metrics Operator is now supported on IBM Power® (ppc64le), IBM Z® (s390x), IBM® LinuxONE (s390x), and ARM (aarch64) architectures. Cost management for OpenShift on IBM Z® is not IFL aware.

We made several changes to the UI that affect how to update your cost management settings. Before, you updated your settings in the Red Hat Hybrid Cloud Console settings page. With this update, cost management has its own settings page.