Capacity Planning Dashboard
The Capacity Planning Dashboard provides insights into current cluster utilization and highlights CPU and memory overcommitment.
To access the dashboard:
- Navigate to Container Cluster Management.
- In the left navigation bar, hover over the Container Cluster Management Dashboard icon.
- Select Capacity Planning.
Users must have an App Owner, Capacity & Performance Engineer or IT Ops Mgr role assigned to access the Capacity Planning Dashboard.
In addition to the Cluster Health by Provider (donut chart), Actionable Insights and the Cluster List table, the dashboard includes three additional tiles that provide an at-a-glance view of capacity and utilization:
- Most utilized clusters: The number of clusters that exceeded the high threshold for CPU, memory or storage in the last 30 days.
- CPU overcommitted clusters: The number of clusters that currently exceed the overcommitted CPU threshold.
- Memory overcommitted clusters: The number of clusters that currently exceed the overcommitted memory threshold.
Thresholds for overcommitment follow:
- CPU utilization:
- Green: 0% - 70% (normal range)
- Amber: 71% - 80% (warning range)
- Red: 81% - 100% (action recommended range)
- Memory utilization:
- Green: 0% - 80% (normal range)
- Amber: 81% - 90% (warning range)
- Red: 91% - 100% (action recommended range)
- Memory utilization:
- Green: 0% - 75% (normal range)
- Amber: 76% - 80% (warning range)
- Red: 81% - 100% (action recommended range)
Each tile contains a View link that navigates to a details table for all affected clusters. Details include critical identifying information such as cluster name, environment, (CPU/memory) utilization and (CPU/memory) request.
Directly below the utilization and overcommitment tiles, the dashboard presents two tables:
- Top 20 Pods by Network I/O: The 20 pods with the highest amount of inbound and outbound network traffic. Click the Pod Name for Utilization details:
- CPU utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Memory Utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Storage utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Network I/O: A graph that provides instantaneous traffic measurement over time in one-day increments.
- Top 20 Pods by Disk I/O: The 20 pods with the highest amount of read and write throughput traffic. Click the Pod Name for Utilization details:
- CPU utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Memory Utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Storage utilization: A graph that provides instantaneous utilization over time in one-day increments.
- Disk I/O: A graph that provides instantaneous disk read and write throughput over time in one-day increments.