Observability into kubernetes cluster

Observability into your Kubernetes cluster

Select Cluster Manager → K8s Cluster

The following three views will give three different levels of visibility into your Kubernetes clusters.

• Cluster View (Default Screen)

• Node View

• Pod View

All Cluster Manager screens have a time filter in the top right corner that can be used to filter by the time intervals.

Fig: Right-click on the cluster/node/pod

From Cluster Manager screens you can do the following operations. It acts as a hybrid screen.

• Add Policies

• Add Label

• View Policies

• View Recommended Policies

You can make use of all these functionalities in all three views of the cluster manager. You have the additional option of View Pods in the Cluster View.

Cluster View

This discusses the Clusters View page and helps you understand the data displayed on the screen.

The Cluster Overview page provides key metrics such as labels, the number of nodes, the number of policies, alerts, etc. of each cluster. Your cluster can reside in any cloud environment of your choice.

Each row represents a cluster. You can further drill down to the Nodes or Pods View page.

Cluster Data

Number of nodes:- Kubernetes runs your workload by placing containers into Pods to run on Nodes. The number shows the available nodes across the entire cluster.

Number of pods:- The number shows available pods across the entire cluster.

Number of Policies:- The number shows the number of active policies across the entire cluster.

Alerts:- Number of alerts across the entire cluster.

Location:- The zone/region in which your cluster (control plane and nodes) are located.

Node View

Left Click on any cluster from the cluster view screen will take you to Node view.

Node Data

Labels:- Number of available labels on specific node.

Number of pods: Number of pods successfully scheduled at a specific node.

Number of Policies: Number of Host polices applied to specific node. Host policies apply to all the nodes selected by their Node Selector.

Alerts: Number of alerts across the specific node.

Location: The zone/region in which your cluster (control plane and nodes) are located.

Pod View

Left click on any node from the node view will take you to Pod view screen. This view will list all the pods in the node along with details.

Pod Data

Workload: Workload column is giving workload identity of the given pod. Workload is identified by the Accuknox workload identification engine. Accuknox will recommend policies based on this workload identification.

Labels: Number of available labels on a specific pod.

Number of containers: Available number of containers inside a pod.

Number of Policies: Number of policies applied to the pod.

Location: The zone/region in which your cluster (control plane and nodes) are located.

Alerts: Number of alerts for the specific node.

View network traffic/Graph view

Network traffic view will give additional flow information between the pods. This network traffic is grouped by namespaces. This view will give much visibility to your workloads.

Allowed traffic is indicated by the green lines and Restricted traffic is indicated by the red lines.

When you right-click on a pod, you can see pod-related details on the right of the screen, and similarly, when you right-click on the flow line you are able to see a connection summary between two pods.

You can also add policies to the connection from the connection summary window.

There is an option to see the entire network traffic across your cluster.

Cluster List view -> click Number of pods -> View Network Traffic

This will give you network traffic for your entire cluster.

Analyzing the flow information of your cluster, you can take decisions on how to secure your workloads at run-time.