Infrastructure for Kubernetes: Clusters

The Kloudfuse Clusters screen gives you a unified view of all Kubernetes clusters, helping you monitor availability, capacity, and performance across multiple regions and cloud providers.

What are Clusters?

Clusters are the highest-level units of a Kubernetes deployment. Each cluster contains a control plane and a set of worker nodes. In Kloudfuse, the Clusters screen provides a unified view of all clusters to help you monitor availability, capacity, and performance across regions and clouds.

Access the Cluster screen

Go to Infrastructure > Kubernetes > Cluster.

Cluster Table Columns

Each column in the table displays key cluster attributes:

1. Cluster – Name of the Kubernetes cluster
2. Status – Cluster availability and health indicators
3. Age – Time since the cluster was created
4. Version – Kubernetes version currently running
5. Nodes – Total number of nodes in the cluster
6. CPU
   • Total – Combined total CPU cores across all nodes
   • Allocable – CPU capacity available for workloads
   • Used – Real-time CPU usage by active pods
7. Memory
   • Total – Combined memory capacity across all nodes (GiB)
   • Allocable – Memory allocable to workloads
   • Used – Real-time memory consumption
8. Pods
   • Total – Total number of running pods in the cluster
   • % Usage – Percentage of allocated pod capacity used

Filters

Kloudfuse provides three categories of filters to scope your results precisely.

• Kubernetes Filter
• Facets Filter
• Additional Filter

Search, group, and sort

Use the following options to refine your search view:

1. Search

Use the Search bar to filter resources such as pods, nodes, clusters, and namespaces using key-value pairs—just like filtering logs in Kloudfuse.

• Auto-suggestions simplify input with live facet and value suggestions.
  • Example:
    1. Type cluster — auto-complete lists matching field names.
    2. Select cluster — a new dropdown appears with current cluster names.
    3. Choose one or more cluster names to apply the filter.

This scoped search helps you quickly locate relevant pods by matching attributes such as:

• cluster

• namespace

• status

• node

• deployment

• and more.

• The search interface dynamically updates based on selected filters and maintains high responsiveness for large clusters.

2. Group By

Group resources—such as Pods, Clusters, Namespaces, or Nodes—by shared attributes to simplify analysis and troubleshooting. Use the icon to toggle between grouped and ungrouped views.

Common groupings include:

• Cluster – See how resources are distributed across clusters.
• Namespace – Organize by logical boundaries within the cluster.
• Node – Identify which node is hosting the resource.
• Status – Group by lifecycle status (Running, Pending, etc.)

Grouping improves visibility into resource alignment, health, and usage patterns.

3. Sort

Use the icon to sort the list by any column in ascending or descending order.

Click the column header to toggle sort direction. Use sorting to identify highest resource usage, longest-running pods, or newest deployments.

Click any row to open the cluster detail pane.

Cluster Detail pane

Each cluster view includes multiple tabs for deeper observability. The node detail pane contains three tabs. These include:

• Logs
• Metrics
• Events

Logs

Use the Logs tab to inspect real-time application and system logs for a selected Kubernetes resource.

• Select specific containers to view logs at a granular level.

Log search supports full-text match, tag selection, and time-range filters.

Metrics

The Metrics tab displays real-time CPU, memory, network, and disk usage for the selected resource.

• Visualize container-level metrics and trends.
• Correlate spikes or anomalies with logs or events.
• Use chart time selectors to view metrics at different intervals (e.g., last 5m, 1h, 24h).

Events

Use the Events tab to view scheduled updates, restarts, warnings, or failed operations for a Kubernetes resource.

• Events include timestamps and source controllers.
• Use this tab to diagnose readiness failures, scaling issues, or resource evictions.
• Combine with Logs and Metrics for faster RCA.