Skip to main content

Resource Leaderboards

The resource leaderboards provide detailed analysis of the highest resource-consuming pods in clusters, helping identify optimization opportunities and performance bottlenecks.

Top CPU Consumers​

Leaderboard Interface​

The resource leaderboard displays the top 10 pods by CPU usage in a comprehensive data grid:

![Figure needed]

Screenshot of resource leaderboard showing top CPU consuming pods

Grid Columns​

Pod Information​

  • Pod Name: Kubernetes pod identifier (clickable for details)
  • Namespace: Pod's namespace location
  • Container Count: Number of containers within the pod
  • Labels: Pod labels with intelligent truncation for display

Resource Metrics​

  • CPU Usage: Current CPU consumption with proper unit formatting
  • Memory Usage: Current memory consumption with automatic unit conversion
  • Age: Duration the pod has been running (e.g., "2d 14h", "3h 42m")

Data Features​

  • Sortable Columns: Click any column header to sort by that metric
  • Real-time Data: Updates with dashboard refresh cycle (30 seconds)
  • Unit Formatting: Automatic conversion to appropriate units (cores, GB, MB)
  • Time Formatting: Human-readable time displays

Data Grid Features​

Interactive Elements​

![Figure needed]

Screenshot showing sortable columns and interactive elements

Column Sorting​

  • Click to Sort: Click any column header to sort data
  • Sort Direction: Toggle between ascending and descending
  • Multi-column: Sort by multiple criteria for detailed analysis
  • Default Sort: Initially sorted by CPU usage (highest first)

Label Management​

  • Smart Truncation: Long label strings intelligently shortened
  • Hover Details: Full label information on mouse hover
  • Key Labels: Most important labels prioritized in display
  • Readable Format: Labels formatted for easy reading

Resource Formatting​

CPU Usage Display​

Format Examples:
- "1.25 cores" (high usage)
- "0.5 cores" (moderate usage)  
- "0.1 cores" (low usage)
- "2.8 cores" (very high usage)

Memory Usage Display​

Format Examples:
- "2.5 GB" (gigabyte range)
- "512 MB" (megabyte range)
- "4.2 GB" (high memory usage)
- "128 MB" (low memory usage)

Age Formatting​

Format Examples:
- "5m" (5 minutes)
- "2h 30m" (2 hours 30 minutes)
- "1d 6h" (1 day 6 hours)
- "3d 14h" (3 days 14 hours)

Analysis Capabilities​

Performance Pattern Identification​

High Resource Consumers​

  • Consistent High Usage: Pods that consistently consume significant resources
  • Resource Spikes: Pods with periodic high resource usage
  • Baseline Consumers: Pods with steady, predictable usage patterns
  • Optimization Candidates: Pods that may benefit from resource tuning

Usage Categories​

High CPU Consumers (over 1 core):
- Database pods
- Compute-intensive applications
- Machine learning workloads
- Data processing jobs

Medium CPU Consumers (0.2-1 core):
- Web applications
- API services
- Monitoring tools
- Standard business applications

Low CPU Consumers (under 0.2 core):
- Logging agents
- Sidecar containers
- Configuration pods
- Lightweight services

Optimization Opportunity Discovery​

Over-provisioned Workloads​

  • High Limits, Low Usage: Pods with resource limits much higher than actual usage
  • Waste Identification: Resources allocated but not utilized
  • Right-sizing Opportunities: Potential for resource limit adjustments
  • Cost Reduction: Opportunities to reduce resource costs

Under-provisioned Workloads​

  • Resource Pressure: Pods approaching or hitting resource limits
  • Performance Impact: Workloads affected by resource constraints
  • Scaling Needs: Pods that may need resource increases
  • Bottleneck Identification: Resource constraints affecting performance

Use Cases​

Daily Operations​

Resource Monitoring​

  1. Quick Assessment: Identify highest resource consumers at a glance
  2. Anomaly Detection: Spot unusual resource consumption patterns
  3. Capacity Planning: Understand which workloads drive resource needs
  4. Performance Tracking: Monitor resource usage trends over time

Issue Investigation​

  1. Performance Problems: Find resource-intensive pods affecting cluster performance
  2. Resource Conflicts: Identify pods competing for resources
  3. Capacity Issues: Understand what's consuming cluster capacity
  4. Troubleshooting: Correlate performance issues with resource usage

Optimization Planning​

Resource Right-sizing​

  1. Usage Analysis: Compare actual usage to resource requests/limits
  2. Efficiency Improvement: Identify opportunities for better resource utilization
  3. Cost Optimization: Find ways to reduce resource costs
  4. Performance Tuning: Optimize resource allocation for better performance

Capacity Management​

  1. Growth Planning: Understand resource growth patterns
  2. Scaling Decisions: Make informed decisions about cluster scaling
  3. Workload Distribution: Plan optimal workload distribution
  4. Resource Allocation: Optimize resource allocation strategies

Performance Optimization​

Workload Analysis​

  1. Resource Patterns: Understand resource consumption patterns
  2. Peak Usage: Identify peak resource usage periods
  3. Baseline Requirements: Establish baseline resource requirements
  4. Scaling Triggers: Determine when workloads need scaling

Cluster Efficiency​

  1. Utilization Rates: Monitor overall cluster utilization efficiency
  2. Resource Balance: Ensure balanced resource usage across nodes
  3. Waste Reduction: Minimize resource waste and over-provisioning
  4. Performance Correlation: Correlate resource usage with application performance

Best Practices​

Regular Monitoring​

  1. Daily Reviews: Check top consumers daily for anomalies
  2. Weekly Analysis: Perform deeper analysis weekly
  3. Trend Tracking: Monitor changes in top consumers over time
  4. Pattern Recognition: Identify recurring resource usage patterns

Optimization Actions​

  1. Resource Requests: Adjust resource requests based on actual usage
  2. Resource Limits: Set appropriate resource limits to prevent resource hogging
  3. Pod Distribution: Ensure even distribution of resource-intensive pods
  4. Scaling Configuration: Configure appropriate horizontal and vertical scaling

Proactive Management​

  1. Alert Thresholds: Set up alerts for unusual resource consumption
  2. Capacity Planning: Use data for proactive capacity planning
  3. Cost Management: Monitor resource costs and optimization opportunities
  4. Performance Optimization: Continuously optimize based on usage patterns

Data Accuracy and Limitations​

Data Sources​

  • Metrics Server: Real-time resource usage from Kubernetes metrics server
  • Update Frequency: Data refreshed every 30 seconds with dashboard
  • Accuracy: High accuracy for operational decision-making
  • Scope: Covers all pods across all namespaces in the cluster

Considerations​

  • Point-in-time: Shows current usage, not historical averages
  • Sampling: Based on most recent metrics collection
  • Variability: Resource usage can vary significantly over time
  • Context: Consider workload characteristics when interpreting data

Integration with Other Monitoring​

Cross-reference with Node Health​

  1. Node Correlation: Correlate top consumers with node resource pressure
  2. Distribution Analysis: Understand how top consumers are distributed across nodes
  3. Capacity Impact: See how top consumers affect overall node capacity
  4. Load Balancing: Ensure top consumers are properly distributed

Analytics Integration​

  1. Namespace Analysis: Cross-reference with namespace resource analytics
  2. Pod Distribution: Correlate with pod distribution patterns
  3. Trend Analysis: Combine with historical trend data
  4. Optimization Planning: Integrate with capacity planning analytics

For additional cluster monitoring capabilities, see the Cluster Management section.