sidebar | permalink | keywords | summary |
---|---|---|---|
sidebar |
kubernetes_landing_page.html |
telegraf, installation, install, kubernetes, pods, clusters, filters, filtering, k8s |
The Kubernetes landing page explained |
The Cloud Insights Kubernetes Explorer is a powerful tool for displaying the overall health and usage of your Kubernetes clusters and allows you to easily drill down into areas of investigation.
Clicking on Dashboards > Kubernetes Explorer opens the Kubernetes Cluster overview page. This overview page contains a wide variety of at-a-glance information.
Note
|
Numbers displayed in blue in the various sections of this and subsequent Kubernetes Explorer pages (for example, node/pod status, namespace counts, etc.) are links to related query pages that show the details behind those numbers. |
The cluster list displays the following usage information for each cluster in your environment:
-
CPU: percentage of total CPU capacity in use
-
Memory: percentage of total memory used
-
Storage: percentage of total storage in use
You can sort the cluster list by any of the following factors:
-
Node & Pod Alert Level (default)
-
Cluster Name
-
Number of Nodes
-
Most Utilized by Compute
-
Least Utilized by Compute
-
Most Utilized by Storage
-
Least Utilized by Storage
Clicking on a Cluster Name will open the detail page for that cluster
To the right of the screen is a list of the top three namespaces utilized in each cluster. Click the All link to see all namespaces for the cluster.
When you are filtering, as you begin typing you are presented with the option to create a wildcard filter based on the current text. Selecting this option will return all results that match the wildcard expression. You can also create expressions using NOT or AND, or you can select the "None" option to filter for null values in the field.
Filters based on wildcards or expressions (e.g. NOT, AND, "None", etc.) display in dark blue in the filter field. Items that you select directly from the list are displayed in light blue.
Kubernetes filters are contextual, meaning for example that if you are on a specific node page, the pod_name filter only lists pods related to that node. Moreover, if you apply a filter for a specific namespace, then the pod_name filter will list only pods on that node and in that namespace.
Note that Wildcard and Expression filtering works with text or lists but not with numerics, dates or booleans.