Skip to main content

Kubernetes Monitoring Metrics

For overall Kubernetes cluster resource utilization, the following metrics are displayed:

Metric NameMetric DescriptionUnit
CPU UsagePercentage of CPU used by all Containers in k8s cluster, formula is container CPU used / Max CPUs.%
Max CPUsTotal CPU allowed for use by all Node nodes in k8s cluster.Cores
Memory UsageMemory usage ratio of all containers in k8s cluster. Formula is total memory used by all Containers in k8s cluster / Max Memory.%
Max MemoryTotal memory allowed for use by all Node nodes in k8s cluster.Bytes
CPU RequestsTotal ratio of Requests CPU configured for all Containers in k8s cluster. Formula is container request cpu / Max CPUs.%
Memory RequestsTotal ratio of Requests memory configured for all Containers in k8s cluster. Formula is container request memory / Max Memory.%
CPU LimitsTotal ratio of limits CPU configured for all Containers in k8s cluster. Formula is container limit cpu / Max CPUs.%
Memory LimitsTotal ratio of limits memory configured for all Containers in k8s cluster. Formula is container limit memory / Max CPUs.%
CPU AvailableRepresents the ratio of CPU resources that can be configured, formula is 1 - CPU Requests.%
Memory AvailableRepresents the ratio of memory resources that can be configured, formula is 1 - Memory Requests.%

For each Node, the following metrics are displayed:

Metric NameMetric DescriptionUnit
StatusRepresents the Node status.Ready/Not Ready
ProblemWhether there are abnormalities such as high CPU usage, high memory usage, full disk space, etc.
CPU UsageCPU usage of all Containers in Node.mCores
CPU RequestsCPU requests amount of all Containers in Node.mCores
CPU LimitsCPU limits amount of all Containers in Node.mCores
Memory UsageMemory usage of all Containers in Node.Bytes
Memory RequestsMemory Requests amount of all Containers in Node.Bytes
Memory LimitsMemory Limits amount of all Containers in Node.Bytes

For each Pod, the following metrics are displayed:

Metric NameMetric DescriptionUnit
Running StatusPod running status.
NamespaceNamespace that the Pod belongs to.
CPU (mCores)CPU usage of all Containers in Pod.mCores
Memory UsageMemory usage of all Containers in Pod.Bytes
Network Throughput (In/Out)Network throughput of all Containers in Pod.bit/s
Disk Throughput (Read/Write)Disk throughput of all Containers in Pod.Bytes/s
Restart CountCount
Tag NameKey value of labels on pod.
Tag ValueValue of labels on pod.
Event TypeEvent type.
Event ReasonEvent reason.
Event TimeEvent time.
Event MessageEvent content.