Delete a cluster machine¶

This section instructs you on how to scale down an existing management, regional, or managed cluster through the Mirantis Container Cloud web UI.

Precautions¶

Before deleting a cluster machine, carefully read the following essential information for a successful machine deletion:

We recommend deleting cluster machines using the Container Cloud web UI or API instead of using the cloud provider tools directly. Otherwise, the cluster deletion or detachment may hang and additional manual steps will be required to clean up machine resources.
An operational managed cluster must contain a minimum of 3 Kubernetes manager nodes and 2 Kubernetes worker nodes. The deployment of the cluster does not start until the minimum number of nodes is created.

A machine with the manager node role is automatically deleted during the cluster deletion.
- Before Container Cloud 2.17.0, to meet the etcd quorum and prevent the deployment failure, deletion of the manager nodes is prohibited.
- Since Container Cloud 2.17.0, deletion of the manager nodes is allowed for non-MOSK-based clusters within the Technology Preview features scope for the purpose of node replacement or recovery.
Consider the following precautions before deleting manager machines from non-MOSK-based clusters:
- Create a new manager machine to replace the deleted one as soon as possible. This is necessary since after machine removal, the cluster has limited capabilities to tolerate faults. Deletion of manager machines is intended only for replacement or recovery of failed nodes.
- You can delete a manager machine only if your cluster has at least two manager machines in the Ready state.
- Do not delete more than one manager machine at once to prevent cluster failure and data loss.
- For the Equinix Metal and bare metal providers, ensure that the machine to delete is not a Ceph Monitor. Otherwise, migrate the Ceph Monitor to keep the odd number quorum of Ceph Monitors after the machine deletion. For details, see Migrate a Ceph Monitor before machine replacement.
- On managed clusters, deletion of a machine assigned to a machine pool without decreasing replicas count of a pool automatically recreates the machine in the pool. Therefore, to delete a machine from a machine pool, first decrease the pool replicas count.
If StackLight in HA mode is enabled and you are going to delete a machine with the StackLight label:
- Make sure that at least 3 machines with the StackLight label will remain after the deletion. Otherwise, add an additional machine with such label before the deletion. After the deletion, perform the additional steps described below.
- Do not delete more than 1 machine with the StackLight label. Since StackLight in HA mode uses local volumes bound to machines, the data from these volumes on the deleted machine will be purged but its replicas remain on other machines. Removal of more than 1 machine can cause data loss.
If the machine being deleted has a prioritized upgrade index and you want to preserve the same upgrade order, manually set the required index to the new node that replaces the deleted one. Otherwise, the new node is automatically set the greatest upgrade index that is prioritized the last. To set the upgrade index, refer to Change the upgrade order of a machine or machine pool.

Delete a machine from a cluster¶

Carefully read the machine deletion precautions.
Log in to the Container Cloud web UI with the m:kaas:namespace@operator or m:kaas:namespace@writer permissions.
Switch to the required project using the Switch Project action icon located on top of the main left-side navigation panel.
In the Clusters tab, click on the required cluster name to open the list of machines running on it.
For the Equinix Metal and bare metal providers, ensure that the machine being deleted is not a Ceph Monitor. If it is, migrate the Ceph Monitor to keep the odd number quorum of Ceph Monitors after the machine deletion. For details, see Migrate a Ceph Monitor before machine replacement.
If the machine is assigned to a machine pool, decrease replicas count of the pool as described in Change replicas count of a machine pool.
Click the More action icon in the last column of the machine you want to delete and select Delete. Confirm the deletion.

Deleting a machine automatically frees up the resources allocated to this machine.

Applicable only to managed clusters. Skip this step if your management cluster is upgraded to Container Cloud 2.17.0.

If StackLight in HA mode is enabled and the deleted machine had the StackLight label, perform the following steps:

Connect to the managed cluster as described in the steps 5-7 in Connect to a Mirantis Container Cloud cluster.

Define the pods in the Pending state:

kubectl get po -n stacklight | grep Pending

Example of system response:

opensearch-master-2             0/1       Pending       0       49s
patroni-12-0                    0/3       Pending       0       51s
patroni-13-0                    0/3       Pending       0       48s
prometheus-alertmanager-1       0/1       Pending       0       47s
prometheus-server-0             0/2       Pending       0       47s

Verify that the reason for the pod Pending state is volume node affinity conflict:

kubectl describe pod <POD_NAME> -n stacklight

Example of system response:

Events:
  Type     Reason            Age    From               Message
  ----     ------            ----   ----               -------
  Warning  FailedScheduling  6m53s  default-scheduler  0/6 nodes are available:
                                                       3 node(s) didn't match node selector,
                                                       3 node(s) had volume node affinity conflict.
  Warning  FailedScheduling  6m53s  default-scheduler  0/6 nodes are available:
                                                       3 node(s) didn't match node selector,
                                                       3 node(s) had volume node affinity conflict.

Obtain the PVC of one of the pods:

kubectl get pod <POD_NAME> -n stacklight -o=jsonpath='{range .spec.volumes[*]}{.persistentVolumeClaim}{"\n"}{end}'

Example of system response:

{"claimName":"opensearch-master-opensearch-master-2"}

Remove the PVC using the obtained name. For example, for opensearch-master-opensearch-master-2:
```
kubectl delete pvc opensearch-master-opensearch-master-2 -n stacklight
```

Delete the pod:

kubectl delete po <POD_NAME> -n stacklight

Verify that a new pod is created and scheduled to the spare node. This may take some time. For example:

kubectl get po opensearch-master-2 -n stacklight
NAME                     READY   STATUS   RESTARTS   AGE
opensearch-master-2   1/1     Running  0          7m1s

Repeat the steps above for the remaining pods in the Pending state.