> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ankra.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Kubernetes Workloads

> Manage Deployments, Pods, StatefulSets, DaemonSets, Jobs, and CronJobs in Ankra

<Note>
  The Workloads section in Ankra gives you complete visibility and control over all running applications in your Kubernetes clusters.
</Note>

## Overview

Workloads are the applications running in your cluster. Ankra provides a unified interface to view, manage, and troubleshoot all workload types:

* **Deployments** - Stateless applications with declarative updates
* **Pods** - The smallest deployable units
* **StatefulSets** - Stateful applications with stable identities
* **DaemonSets** - Pods that run on every node
* **ReplicaSets** - Maintain replica pod counts
* **Jobs** - Run-to-completion tasks
* **CronJobs** - Scheduled jobs
* **Horizontal Pod Autoscalers (HPAs)** - Automatic scaling

***

## Accessing Workloads

Navigate to your cluster and click **Kubernetes** in the sidebar. The Workloads section includes:

| Resource     | Path                                    |
| ------------ | --------------------------------------- |
| Deployments  | Kubernetes → Deployments                |
| Pods         | Kubernetes → Pods                       |
| StatefulSets | Kubernetes → StatefulSets               |
| DaemonSets   | Kubernetes → DaemonSets                 |
| ReplicaSets  | Kubernetes → ReplicaSets                |
| Jobs         | Kubernetes → Jobs                       |
| CronJobs     | Kubernetes → CronJobs                   |
| HPAs         | Kubernetes → Horizontal Pod Autoscalers |

Or use the Command Palette (`⌘+K`) to jump directly to any resource type.

***

## Deployments

Deployments manage ReplicaSets and provide declarative updates for Pods.

### Viewing Deployments

The Deployments list shows:

| Column     | Description                       |
| ---------- | --------------------------------- |
| Name       | Deployment name                   |
| Namespace  | Kubernetes namespace              |
| Ready      | Ready replicas / Desired replicas |
| Up-to-date | Pods running the latest spec      |
| Available  | Pods available for traffic        |
| Age        | Time since creation               |

### Deployment Details

Click a deployment to view:

* **Status** - Current rollout state and conditions
* **Replicas** - Desired, current, ready, and available counts
* **Strategy** - Rolling update or recreate
* **Pod Template** - Container specs, resources, environment
* **Events** - Recent Kubernetes events
* **Managed Pods** - List of pods owned by this deployment

### Actions

* **Scale** - Adjust replica count
* **Restart** - Trigger a rolling restart
* **View YAML** - See the full resource definition
* **Troubleshoot** - AI-assisted diagnosis
* **Delete** - Remove the deployment

***

## Pods

Pods are groups of containers that share storage and network.

### Viewing Pods

The Pods list shows:

| Column    | Description                         |
| --------- | ----------------------------------- |
| Name      | Pod name                            |
| Namespace | Kubernetes namespace                |
| Ready     | Ready containers / Total containers |
| Status    | Running, Pending, Failed, etc.      |
| Restarts  | Container restart count             |
| Age       | Time since creation                 |
| Node      | Node the pod is scheduled on        |

### Pod Details

Click a pod to view:

* **Containers** - List of containers with status, image, and resources
* **Conditions** - PodScheduled, ContainersReady, Ready
* **Events** - Recent events (scheduling, pulling, started, etc.)
* **Labels & Annotations** - Metadata
* **Volumes** - Mounted volumes and claims

### Pod Logs

View real-time and historical logs:

1. Click on a pod
2. Select the **Logs** tab
3. Choose the container (for multi-container pods)
4. Toggle **Follow** for real-time streaming
5. Search within logs using the filter

### Actions

* **View Logs** - Stream container logs
* **View YAML** - Full pod specification
* **Troubleshoot** - AI analysis of pod issues
* **Delete** - Remove the pod (will be recreated by controllers)

***

## StatefulSets

StatefulSets manage pods with persistent identities and stable storage.

### Key Features

* **Stable Pod Names** - Pods have predictable names (app-0, app-1, etc.)
* **Persistent Storage** - Volume claims are retained across restarts
* **Ordered Operations** - Pods are created/deleted in order

### Viewing StatefulSets

The list shows replica status, update status, and age. Details include:

* Pod management policy
* Update strategy
* Volume claim templates
* Managed pods

***

## DaemonSets

DaemonSets ensure a pod runs on every (or selected) node.

### Use Cases

* Log collectors (Fluentd, Filebeat)
* Node monitoring (Prometheus node exporter)
* Network plugins (CNI, kube-proxy)
* Storage plugins (CSI drivers)

### Viewing DaemonSets

The list shows:

| Column     | Description                   |
| ---------- | ----------------------------- |
| Desired    | Pods that should be scheduled |
| Current    | Currently running pods        |
| Ready      | Pods ready to serve           |
| Up-to-date | Pods with latest spec         |
| Available  | Pods available                |

***

## Jobs & CronJobs

### Jobs

Jobs create pods that run to completion:

* **Completions** - How many successful completions are needed
* **Parallelism** - How many pods can run concurrently
* **Status** - Active, succeeded, failed counts

### CronJobs

CronJobs schedule Jobs on a time-based schedule:

* **Schedule** - Cron expression (e.g., `0 */6 * * *`)
* **Last Schedule** - When it last ran
* **Active** - Currently running jobs
* **Suspend** - Whether scheduling is paused

***

## Horizontal Pod Autoscalers

HPAs automatically scale workloads based on metrics.

### Viewing HPAs

| Column    | Description                   |
| --------- | ----------------------------- |
| Name      | HPA name                      |
| Reference | Target deployment/statefulset |
| Min/Max   | Replica bounds                |
| Replicas  | Current replica count         |
| Metrics   | CPU/Memory utilization        |

### HPA Details

* Target metrics and current values
* Scaling history
* Conditions and events

***

## Bulk Delete Resources

You can select and delete multiple Kubernetes resources at once across all resource types.

### How to Bulk Delete

<Steps>
  <Step title="Navigate to a Resource List">
    Go to any Kubernetes resource list (Deployments, Pods, StatefulSets, Services, ConfigMaps, etc.).
  </Step>

  <Step title="Select Resources">
    Click the checkbox next to each resource you want to delete. Use the top checkbox to select all visible resources.
  </Step>

  <Step title="Click Delete">
    A toolbar appears showing the number of selected items. Click **Delete Selected** to begin.
  </Step>

  <Step title="Confirm Deletion">
    Review the list of resources to be deleted in the confirmation dialog. Confirm to proceed.
  </Step>
</Steps>

Resources are deleted sequentially with progress tracking. If any individual deletion fails, the error is reported per item and the remaining deletions continue.

### Supported Resource Types

Bulk delete is available across all Kubernetes resource types in Ankra, including:

* Workloads: Deployments, Pods, StatefulSets, DaemonSets, ReplicaSets, Jobs, CronJobs
* Networking: Services, Ingresses, Network Policies, Endpoints
* Configuration: ConfigMaps, Secrets
* Storage: Persistent Volume Claims, Persistent Volumes, Storage Classes
* RBAC: Roles, RoleBindings, ClusterRoles, ClusterRoleBindings, ServiceAccounts

***

## Common Tasks

### Troubleshooting a Failing Pod

1. Navigate to **Pods** and find the failing pod
2. Check the **Status** column for error indicators
3. Click the pod and review:
   * **Events** for scheduling or pull errors
   * **Logs** for application errors
   * **Conditions** for readiness issues
4. Click **Troubleshoot** for AI-assisted diagnosis

### Scaling a Deployment

1. Navigate to **Deployments**
2. Click on the deployment to scale
3. Click **Scale** and enter the new replica count
4. Confirm the change

### Restarting Pods

1. Navigate to **Deployments**
2. Click on the deployment
3. Click **Restart** to trigger a rolling restart
4. Monitor the rollout in **Events**

***

## Nodes

The **Nodes** view under **Kubernetes → Nodes** shows all nodes in your cluster with their status, roles, resource usage, and scheduling state.

### Node List

The node list displays:

| Column       | Description                           |
| ------------ | ------------------------------------- |
| Name         | Node hostname                         |
| Status       | Ready / NotReady and scheduling state |
| Roles        | Control plane, worker, etcd           |
| Version      | Kubelet version                       |
| Age          | Time since node joined the cluster    |
| CPU / Memory | Allocatable resources                 |

Nodes that are cordoned show a **SchedulingDisabled** label next to their status.

### Node Actions

#### Cordon & Uncordon

Cordoning a node marks it as unschedulable — no new pods will be placed on it, but existing pods continue running. This is the first step when preparing to move workloads off a node.

* **Single node**: Click a node to open its detail view, then click **Cordon** or **Uncordon** in the header.
* **Bulk action**: Select multiple nodes using the checkboxes, then click **Cordon** or **Uncordon** in the floating action bar. The action bar dynamically shows the relevant buttons based on the scheduling state of selected nodes.

#### Drain

Draining a node cordons it and then evicts all running pods (ignoring DaemonSets and deleting emptyDir data). This is equivalent to:

```bash theme={null}
kubectl drain <node-name> --ignore-daemonsets --delete-emptydir-data
```

* **Single node**: Click a node, then click **Drain** in the header.
* **Bulk action**: Select one or more nodes, then click **Drain** in the floating action bar.

The drain operation shows real-time progress: cordoning, fetching pods, and evicting each pod sequentially. Failed evictions are reported per pod while the remaining evictions continue.

### Moving Workloads to New Node Groups

To migrate workloads from old nodes to a new node group:

<Steps>
  <Step title="Create a new node group">
    Add a new node group in **Cluster Settings → Nodes** with your desired instance type, labels, and taints.
  </Step>

  <Step title="Cordon old nodes">
    In **Kubernetes → Nodes**, select the old nodes and click **Cordon** to prevent new scheduling.
  </Step>

  <Step title="Drain old nodes">
    With the same nodes selected, click **Drain** to evict all running pods. Pods managed by Deployments or StatefulSets will be automatically rescheduled onto the new node group.
  </Step>

  <Step title="Remove old node group">
    Once all pods have migrated, delete the old node group from **Cluster Settings → Nodes**.
  </Step>
</Steps>

***

## Tips

<Tip>
  **Use Namespace Filters:** Filter by namespace to focus on specific applications or environments.
</Tip>

<Tip>
  **Watch Events:** Kubernetes events often explain why pods are failing to start.
</Tip>

<Tip>
  **Check Resource Limits:** Many pod failures are due to insufficient CPU or memory limits.
</Tip>

***

Still have questions? [Join our Slack community](https://join.slack.com/t/ankra-community/shared_invite/zt-3a5rem8f8-cUho4epX2MoLT83bFf~VSA) and we'll help out.