Move pods from default system pool to user pool

Question

Move pods from default system pool to user pool

JMN-2253 636

Hi there,

I'm new one to this infinite game, and looking for your help.

Did much research on how to move pods from the default system pool to the user pool, and all the results recommend deleting the containers and recreating them.

How can I achieve it with 0 downtime?

What is the command to build a new one on the node pool called linuxdev while keeping the internal IP?

What is the command to delete the old one?

Thanks

Accepted answer

0 additional answers

Your answer

Answer 1

Prrudram-MSFT 28,366 Microsoft Employee Moderator

@JMN-2253

Hello! I'd be happy to help you with your questions.

To move pods from the default system pool to a user pool, you can use the kubectl cordon command to mark the node as unschedulable, then use the kubectl drain command to evict the pods from the node. Once the pods are evicted, you can delete the node pool and create a new one with the desired configuration.

To minimize downtime, you can create the new node pool before evicting the pods, and then use the kubectl uncordon command to mark the new node pool as schedulable once it is ready.

To build a new container on a node pool called linuxdev while keeping the internal IP, you can use the kubectl apply command with a YAML file that specifies the new pod configuration. Here's an example YAML file:

apiVersion: v1
kind: Pod
metadata:
  name: my-pod
spec:
  nodeName: linuxdev
  containers:
  - name: my-container
    image: my-image

You can save this YAML file to a file called my-pod.yaml, and then apply it to your cluster with the following command:

kubectl apply -f my-pod.yaml

To delete the old container, you can use the kubectl delete command with the name of the pod**1**. Here's an example command:

kubectl delete pod <pod-name>

This command will delete the specified pod, which will cause it to be recreated on a different node if necessary.

If the answer provided has helped, please "Accept Answer" and click on Yes for "was this answer helpful" wherever the information provided helps you, this can be beneficial to other community members.

JMN-2253 636 Reputation points

2023-05-12T15:16:09.4833333+00:00

Hi @Prrudram-MSFT I got this error: node/aks-linuxdevpods-15929581-vmss000001 already cordoned

error: unable to drain node "aks-linuxdevpods-15929581-vmss000001" due to error: cannot delete DaemonSet-managed Pods

I just want to know what that means?
Prrudram-MSFT 28,366 Reputation points Microsoft Employee Moderator

2023-05-13T19:19:46.0366667+00:00
JMN-2253

The error message "unable to drain node due to error: cannot delete DaemonSet-managed Pods" means that there are one or more DaemonSet-managed pods running on the node that you are trying to drain. DaemonSets are a type of Kubernetes workload that ensures that a copy of a pod is running on every node in the cluster. When you try to drain a node that has DaemonSet-managed pods running on it, Kubernetes will not allow you to delete those pods because they are required to be running on every node.

In your case, the error message "node already cordoned" means that the node has already been cordoned, which is the first step in the node draining process. Cordoning a node means that Kubernetes will stop scheduling new pods on the node, but it will not evict any existing pods. This is done to ensure that the node is not overloaded with new pods while you are trying to drain it.

To resolve the error, you can use the --ignore-daemonsets flag with the kubectl drain command. This flag will ignore any DaemonSet-managed pods running on the node and allow you to drain the node. Here's an example command:

kubectl drain <node-name> --ignore-daemonsets

Replace <node-name> with the name of the node that you want to drain. This command will gracefully evict the non-DaemonSet pods from the node and reschedule them on other nodes in the cluster. The DaemonSet-managed pods will remain running on the node. Once the non-DaemonSet pods have been rescheduled, you can safely delete the node.
JMN-2253 636 Reputation points

2023-05-13T20:59:39.47+00:00

Thanks a lot

Share via

Move pods from default system pool to user pool

0 additional answers

Your answer