Viewing your Kubernetes resources
You can view a snapshot of the health and status of your entire Kubernetes resources. In the following example image, there are four healthy clusters housing 11 nodes and 2286 pods.
You can view the following components in the Kubernetes Summary:
Icon | Description |
|---|---|
| Clusters: The total number of Kubernetes clusters you are monitoring, along with how many are healthy or unhealthy. |
| Nodes: The number of nodes, which includes servers or virtual machines in your clusters, and how many are healthy or unhealthy. |
| Pods: The total number of pods running across all clusters, and how many are unhealthy. |
| Deployments: Deployments manage the rollout and scaling of your applications. If some are unhealthy, parts of your app may not be running as expected. |
| DaemonSets: DaemonSets ensure that a copy of a pod runs on every node. If a DaemonSet is unhealthy, some nodes might be missing important background services like logging or monitoring agents. |
| StatefulSets: StatefulSets are used for applications that need stable network identities or persistent storage, like databases. Unhealthy StatefulSets can lead to data loss or unavailable services. |
| Persistent Volumes (PVs): PVs provide storage for your applications. If any are unhealthy, your apps might not be able to read or write data, leading to failures. |
| Persistent Volume Claims (PVCs): PVCs are requests for storage by your applications. Unhealthy PVCs show that some apps can’t access the storage they need, which can cause errors or downtime. |
| Services: Services reflect your applications to the network. If any are unhealthy, users or other apps might not be able to connect, causing disruptions. |
Clusters
The cluster window provides a high-level summary of all Kubernetes clusters in a tabular format, offering essential metrics and actionable insights. Each row represents a cluster, with columns displaying various metrics such as health status, node and pod counts, and resource utilization.
You can quickly identify clusters needing attention and the health impacts of all downstream resources in the clusters window.
You can use Filter by to select clusters based on criteria such as environment or project.
Hover over the Monitoring Health column to view the system status of a particular cluster.
Click the cluster names to access a detailed view for each cluster in Container Observability.
You can customize your metric table with the Edit Columns.
Click the Healthy/Unhealthy under the Health column to view the particular cluster health analysis. You can view the Cluster health, Health Check Summary, and Health Check Details.
Nodes Overview
The Nodes Overview provides a centralized, detailed, and interactive view of your Kubernetes nodes. It enables you to efficiently monitor node health, troubleshoot issues, and optimize resource allocation across clusters.
Filter nodes by specific clusters using the Filter By option.
You can quickly assess overall node health with metrics like Total Nodes, Ready Nodes, NotReady Nodes, and Unknown Nodes.
Expand clusters to view node-level metrics such as Alerts, Status, Condition, Pod Status, Age, and Labels.
Click the node names to access a detailed view for each node in Container Observability.
Pod Overview
In Pod Overview, you can track the total number of pods, CPU/memory usage, restarts, and network activity.
To customize your view, select the cluster you want to examine in the Filter By field. You can also search for the cluster/pod in the search bar below the graph.
In the graph, check the health of the pods. Failed, Pending, and Unknown pods appear in the graph. You can set the graph to show the top 5, 10 or 15 pods with the most issues. If all pods are healthy, then the graph is empty.
Along the top of the graph, you can use the snapshot table to view number totals for your running and succeeded pods as well as the total number of pods in the specified cluster.
When finished, you can collapse the graph and examine the Cluster Name table.
Use the search bar to search for the one you want to view. For each cluster, the table lists the total number of active alerts, pods, running pods, and so forth. To show or hide these columns, click Edit Columns.
Expand your cluster to view the pods in the table.
The Pods table provides you with a snapshot of the pods in the selected cluster so you can investigate any major issues. To view more details about a specific pod, click its link Configuration Snapshot Details of Pod in a Node to open it in your Virtana IO instance.
The table summarizes key information, such as the name of the pod, its namespace location, any active alerts, and its status, among other data. To show or hide these columns, click Edit Columns.
Deployments Overview
It is designed for both high-level summaries and granular insights. This window empowers you to quickly assess deployment health, identify issues, and take informed action. With interactive features, customizable views, and robust export capabilities, the Deployment Overview window streamlines your workflow and enhances operational efficiency.
You can view the graph of deployments with unavailable replicas by cluster. Hover over the graph to view the total number of unavailable replicas by each particular cluster. In the top right corner of the graph, you can view additional information such as available, unavailable, updated, and desired deployments by cluster.
Click the cluster names to expand or hide detailed deployment lists within each cluster to view the details, such as Alerts, Status, Ready Replicas, Namespace, Update Policy, Age, and Images.
Click the deployment names to access a detailed view for deployment in Container Observability.
DaemonSets Overview
It provides both high-level summaries and detailed, interactive insights, empowering you to maintain cluster health and quickly address any issues that arise.
You can view the graph of DaemonSets with unavailable replicas by cluster. In the top right corner of the graph, you can view additional information such as available, unavailable, updated, and desired DemonSets by cluster.
Click the cluster names to expand or hide detailed DemonSet lists within each cluster to view the details, such as Alerts, Status, Ready Pods, Namespace, and Labels.
Click the DemonSet names to access a detailed view of DemonSet in Container Observability.
StatefulSets Overview
In the StatefulSets Overview, you gain a centralized, actionable view of your stateful workloads, making it easier to maintain reliability and performance across your Kubernetes clusters in the Virtana Platform.
A graph shows you the number of StatefulSets with Not Ready Replicas in each cluster. In the top right, you’ll see summary metrics for all StatefulSets, such as Ready, Not Ready, Updated, and Desired.
Click the cluster names to expand or hide detailed StatefulSet lists within each cluster to view the details, such as StatefulSet Name, Alerts, Status, Ready Replicas, Namespace, Description and Labels.
Click the StatefulSet names to access a detailed view of the StatefulSet in Container Observability.
Persistent Volume (PVs) Overview
The Persistent Volume Overview provides a comprehensive and interactive summary of all persistent volumes (PVs), making it easier to monitor, troubleshoot, and optimize persistent storage across your Kubernetes clusters in the Virtana Platform.
You can view a graph that shows the number of PVs with failed/unavailable status in each cluster. In the top right, you’ll see summary metrics for all PVs, such as total PVs, Bound, Available, Pending, Failed, Capacity, and Unbound by cluster.
Click the cluster names to expand or hide the detailed PVs list within each cluster to view the details, such as Persistent Volume Name, Alerts, Phase, Volume Mode, Storage Class, Access Mode, Reclaim Policy, and Capacity.
Click the Persistent Volume names to access a detailed view of the PVs in Container Observability.
Persistent Volume Claim (PVCs) Overview
This section empowers you to view your storage claims, monitor, troubleshoot, and optimize persistent storage usage across your Kubernetes clusters.
A visual graph displays the status of Persistent Volume Claims by cluster. By default, it shows the “Top 5” clusters with the most activity or issues. In the top right of the graph, you can view Total PVCs, Bound (Number of PVCs currently bound to a volume), Pending, and Lost (PVCs in a lost state) in a cluster.
Click the cluster names to expand or hide the detailed PVCs list within each cluster to view the details, such as PVC Name, Alerts, Phase, Persistent Volume, Storage Class, Desired Access Mode, Requested Capacity, and Capacity.
Click the PVC names to access a detailed view of the PVCs in Container Observability.
Service Overview
You can view your Kubernetes services, monitor, troubleshoot, and optimize service performance across your clusters in the Service Overview window.
View the error counts for the top clusters in the graph titled Error Count of Services by Clusters. Customize the graph by selecting Top 5, Top 10, or Top 15 clusters for the time range of 1 hour (1H), 8 hours (8H), or 24 hours (24H).
In the metrics below the graph, you can view the Cluster Name, Alerts, Services, Node Ports, Load Balancers, Max Latency, Error Count, and Max Request.
Click the cluster names to expand or hide the detailed Services list within each cluster to view the metric details, such as Service Name, Alerts, Namespace, IP Address, Type, Port, Max Latency, Error Count, and Max Request.
Click the Service names to access a detailed view of the services in Container Observability.
If you want to share this report or save it locally, click Download to save the data as a PDF or CSV document.
Alerts Trend Analysis
The Alert Trends section in the Cluster Overview Dashboard helps you keep track of all the alerts or anomalies happening in your Kubernetes environment over time. This gives you a clear, visual summary of how many alerts have occurred, how severe they are, and when they happened.
You can track all alerts over time, broken down by the following severity:
Critical: Alert signals the most severe problems in your environment. These usually indicate outages, failed services, or issues that could cause immediate downtime or data loss. It requires immediate action.
Major: Point to significant issues that could affect performance or reliability, but may not cause an immediate outage. These might include resource limits being reached, degraded services, or repeated failures. It is important, but not urgent.
Warning: Warning alerts highlight potential issues or early signs of trouble. These could be things like rising resource usage, minor errors, or configuration mismatches. Impacted resources must be tracked.
Info: Info alerts are informational messages. They let you know about normal events, changes, or updates in your environment, such as successful deployments or routine status changes. This information is for your awareness.
On the top right of the Alerts trend window, you can see the number of impacted resources for each severity.
Hover over the graph to view the severity of a particular point in the graph.
To view in detail, click the alert description below the graph. See Alerts Detail.
Click Open Alerts List to see all alerts and filter them by severity or cluster. See Alerts to view the alerts section in detail.