Part 2: Visualization

What You Will Do

This is Part 2 of a multi-part, self-paced quick start exercise. In this part, you will visualize cost metrics for the cluster that has been set up with cost management.

Important
Cost metrics are collected hourly from the cluster. A wait time of 1 hour is needed for the data to reflect in the dashboard after cost management configuration.

Visualize Cost Metrics

• Navigate to Home -> Dashboards -> Cost Management
• By default, Cost Overview for All will be displayed. You can select specific projects and/or clusters by navigating to the respective tabs
• You can view the summary of total, utilized, idle and unallocated costs along with cost efficiency
• The data is aggregated for a specific time window selected (24 hours/7 days/14 days/30 days)

Cost Trend

• Based on the time window selected, the cost trend for that time window is shown as a time series graph
• For details on granular contribution of costs, hover over the graph
• Costs for each cloud provider can also be viewed separately

Top 5

• A summary of top 5 namespaces, top 5 clusters and top 5 applications are displayed by default
• Click on the ellipses at the top-right corner of each widget to view the complete list of namespaces/clusters/cards ordered by spend

All tab

All tab is an aggregation of all costs in the organization (Across all projects, all clusters). Access to All tab is restricted only to Org Admins.

Project Tab

Project tab is useful to aggregate costs across a project. The summary cards, cost trend and top 5 widgets show the aggregation of costs for the selected project.

Cluster Tab

Cluster tab can be used to view costs for a specific cluster. The summary cards, cost trend and top 5 widgets show the aggregation of costs for the selected cluster.

Recap

Congratulations! You have successfully visualized the cost metrics for the cluster provisioned with cost management. For more accurate cost computations (including negotiated pricing, SPOT instance pricing), custom profile needs to be created and deployed on the cluster.