Sysdig Documentation

Nodes Overview

A node refers to a worker machine in Kubernetes. A physical machine or VM can represent a node.

The Nodes Overview page provides key metrics indicating the health, capacity, and compliance of each node in your cluster.

nodes_overview.png

Scope

The scope of Nodes overview is dictated by two Kubernetes objects: Cluster and Nodes. Both the AND and OR operators are used to determining the scope: Clusters AND Nodes ( a OR b OR c).

nodes_overview_scope.png

Understanding Nodes Overview Metrics

Description

Color Scheme

Example

Metrics

Ready Status

Represented as Node Ready . Shows the latest value of kubernetes.node.ready expressed as a percentage.

If the latest value is 1, the status is Yes.

if the latest value less than 1, status is No.

Red: The nodes are not ready.

Green: The nodes are ready.

The 6 blocks in the Nodes Ready Status represent 6 samples indicating the past status in a selected time window. For a selected period, say an hour, nodes are split into 6 different samples, each one indicating the status for 10 minutes.

The Metric values are visualised based on the time range selected.

80% indicates the latest value of the metrics across the nodes is 0.80.

The value returned by kubernetes.node.ready.

Time aggregation is Minimum.

Group aggregation is Average.

Pod Gauge

Represented as Pods Ready vs Allocatable. This is the ratio between the available and allocatable number of pods for the cluster across deployments, statefulSets, and DaemonSets.

Pods should be 0-80% available.

Red: The number of pods that are ready are less than the allocatable number. The ratio between ready and allocatable pods is greater than 95-100%.

Yellow: The ratio of available and desired pods is between 80-95%.

Green: The ratio of available and desired pods is between 0-80%.

should be 95-100% available.

94% indicates that out of 66 sample pods, approximately 62 pods are available. Therefore indicating in red.

The ratio of kubernetes.pod.status.ready and kubernetes.node.allocatable.pods.

Time aggregation is Average.

Group aggregation is Sum.

CPU Gauge

Represented as CPU Requested vs Allocated. This is the ratio of the number of CPU cores requested and the number of CPU cores allocated by each deployment.

CPU gauge dictates the maximum amount of CPU that your container can use independent of contention on the node. The CPU request represents a minimum amount of CPU that a container can consume. If a container attempts to use more than the specified limit, the system throttles the container.

Red: The ratio between used and requested CPU cores is between 95-100%.

Yellow: The ratio of used and requested CPU cores is between 80 -95%.

Green: The ratio of used and requested CPU cores is less than 80%.

12% indicates out of 50 CPU cores that can be allocated only 6 CPU cores are requested.

The ratio of kubernetes.pod.resourceRequests.cpuCores and kubernetes.node.allocatable.cpuCores.

Time aggregation is Average.

Group aggregation is Sum.

Memory Gauge

Represented as Memory Requested vs Allocated. This is the ratio between the total number of memory requested and the total number of memory a in bytes by each deployment.

Red: The ratio of requested and allocatable memory in bytes is between 95-100%.

Yellow: The ratio of requested and allocatable memory in bytes is between 80%-95%.

Green: The ratio of requested and allocatable memory in bytes is between 0-80%.

2% indicates out of 190 Gib memory that can be allocated, only 3.8 Gib is requested.

The ratio of kubernetes.pod.resourceRequests.memBytes and kubernetes.node.allocatable.memBytes.

Time aggregation is Average.

Group aggregation is Sum.

Network I/O

A sparkline graph

The value returned by net.bytes.total.

Time aggregation is Average.

Group aggregation is Average.

Events

Shows the severity level and the number of events for each type of severity—High, Low, Medium, and Info—in that order.

Red: The number of events that are in High severity state.

Orange: The number of events that are in Medium severity state.

Green: The number of events that are in Low severity state.

Blue: The number of events that are Info.

Drill-Down Features

Kubernetes Nodes Overview

Takes to the Explore page for Nodes Overview. The Nodes Overview highlights the Node’s availability and health as well as resource utilization.

Kubernetes Compliance Report

Record of Kubernetes benchmark test.

Red: The value is between 0-50%.

Yellow: The value is between 50%-80%.

Green: The value is between 80-100%.

The value returned by compliance.k8s-bench.pass_pct.

Time aggregation is Average.

Group aggregation is Average.

Docker Compliance Report

Record of Docker benchmark test.

Red: The value is between 0-50%.

Yellow: The value is between 50%-80%.

Green: The value is between 80-100%.

The value returned by compliance.docker-bench.pass_pct.

Time aggregation is Average.

Group aggregation is Average.