Manage nodes in a cluster

In Cortex XSOAR 8.10 On-prem, manage cluster nodes from the textual UI.

If you deployed your Cortex XSOAR environment starting with three nodes, use the textual UI menu in your VM to add a node, taint a node, remove a node, drain a node, and uncordon a node.

Important

• If you deployed your Cortex XSOAR environment as a standalone (single node), you cannot add nodes to it and switch to a cluster.
• If you are using the OVA image format to deploy a new node, ensure it is deployed from an OVA image that matches the cluster's current Cortex XSOAR version and build. Adding a node created from a different version's OVA image is not supported and will cause the Add Node operation to fail.
• A Kubernetes cluster consists of a control plane and one or more worker nodes. For Cortex XSOAR, in standalone (one VM), the VM acts as both control plane and as a worker node. In multi-node clusters, the first three nodes act as both control plane and as worker nodes, and any additional node added acts as a worker node.

  If you remove one of the original three nodes in the cluster (one of the control planes), you cannot perform actions such as upgrade or scaling up. When you add a new node, Cortex XSOAR automatically assigns the new node as a control plane with the same IP address as the node that was removed.
• You need to set the host again and reestablish trust between all the nodes if you want to add more nodes to the cluster after completing installation.

Add a node

Add a node to a cluster to increase its capacity and improve robustness for better load distribution. For example, by adding more nodes, playbook runs can be divided among them. This reduces the strain on each node and lowers the risk of system overload, while the workload capacity remains the same within the cluster.

Prerequisite

If using the OVA image format:

• Check the current cluster version in the Cortex XSOAR tenant by clicking your user profile → About, and note the Installer Version (for example, master-8.10.0-8.10.0.16-e159785c).
• Deploy the new node using the exact version and build of the current cluster. If the cluster was upgraded since installation, use the OVA image matching the upgraded (current) version; do not use an older OVA image.

1. From the textual UI menu in your VM, select Cluster Administration.
2. Set the host again and reestablish trust between all nodes in the cluster, including the new node (see Task 5. Establish trust between all nodes in a cluster).
3. Select Add Node.
4. Enter the IP Address and click Add.

Taint a node

Tainting a node marks the node as out of service for internal K8s functions. Taint a node to stop applications from running on it.

1. From the textual UI menu in your VM, select Cluster Administration.
2. Select the IP address of the node you want to taint.
3. Select Taint.

   In the list of nodes in the Cluster Administration menu, the node IP will display as Ready,SchedulingDisabled.

Remove a node

Remove a node from a cluster to reduce resources, perform maintenance, or decommission the node, ensuring the cluster operates efficiently without unnecessary or malfunctioning components.

Important

The Remove option in the TUI is only available when the node is in the NotReady state. Draining a node alone leaves it in Ready,SchedulingDisabled and the Remove option will not appear. Shut down the node first and wait until it reports NotReady; the Remove option will then be available in the node action menu.

1. From the textual UI menu in your VM, select Cluster Administration.
2. Drain the node.

   1. Select the IP address of the node you want to drain.
   2. Select Drain.

      In the list of nodes in the Cluster Administration menu, the node IP will display as Ready,SchedulingDisabled.
3. Shut down the node.

   1. Access the VM hosting the node you just drained and perform a shutdown.
   2. Wait until the node reports NotReady in the cluster list.
4. Remove the node.

   1. Select the IP address of the node you want to remove.
   2. Select Remove.

      In the list of nodes in the Cluster Administration menu, the node IP will display as Ready.

Drain a node

Draining a node pauses the node activity in the cluster and marks it as unschedulable. Draining a node safely removes workloads from it, ensuring that running applications are gracefully terminated or moved to other nodes without disrupting service availability before you perform maintenance on the node.

1. From the textual UI menu in your VM, select Cluster Administration.
2. Select the IP address of the node you want to drain.
3. Select Drain.

   In the list of nodes in the Cluster Administration menu, the node IP will display as Ready,SchedulingDisabled.

Uncordon a node

Uncordon a node in a cluster to make it available again for scheduling new workloads, for example after maintenance or troubleshooting is complete.

1. From the textual UI menu in your VM, select Cluster Administration.
2. Select the IP address of the node you want to uncordon.
3. Select Uncordon.

   In the list of nodes in the Cluster Administration menu, the node IP will display as Ready.