ContainerLowCpuUtilization #

Container CPU utilization is under 20% for 1 week. Consider reducing the allocated CPU.

Alert Rule

alert: ContainerLowCpuUtilization
annotations:
  description: |-
    Container CPU utilization is under 20% for 1 week. Consider reducing the allocated CPU.
      VALUE = {{ $value }}
      LABELS = {{ $labels }}    
  runbook: https://srerun.github.io/prometheus-alerts/runbooks/google-cadvisor/containerlowcpuutilization/
  summary: Container Low CPU utilization (instance {{ $labels.instance }})
expr: (sum(rate(container_cpu_usage_seconds_total{container!=&#34;&#34;}[5m])) by (pod, container)
  / sum(container_spec_cpu_quota{container!=&#34;&#34;}/container_spec_cpu_period{container!=&#34;&#34;})
  by (pod, container) * 100) &lt; 20
for: 7d
labels:
  severity: info

Here is the runbook for the Prometheus alert rule ContainerLowCpuUtilization:

Meaning #

The ContainerLowCpuUtilization alert is triggered when a container’s CPU utilization falls below 20% for a period of 7 days. This alert is classified as informational and is not critical, but it may indicate inefficient resource allocation.

Impact #

Low CPU utilization can lead to wasted resources and inefficient use of cluster capacity. If left unchecked, it can result in:

Underutilized nodes and clusters
Inefficient resource allocation
Higher costs due to unused resources
Potential performance issues due to overprovisioning

Diagnosis #

To diagnose the issue, follow these steps:

Identify the affected container and pod using the instance label provided in the alert.
Check the container’s CPU usage over the past 7 days using a tool like Prometheus or Grafana.
Verify that the container is not experiencing any performance issues or errors that could be contributing to the low CPU utilization.
Review the container’s resource allocation and configuration to ensure it is properly sized for its workload.

Mitigation #

To mitigate the issue, consider the following steps:

Reduce the allocated CPU resources for the container to match its actual usage.
Right-size the container’s resource allocation based on its workload and performance requirements.
Consider consolidating underutilized containers onto fewer nodes to optimize resource usage.
Implement automated resource scaling and optimization tools to detect and adjust resource allocation in real-time.

Note: This runbook provides general guidance and may need to be tailored to your specific use case and infrastructure.