Schedule Kubernetes Pods to Nodes

In this article, we will see that how to schedule the pod to run in a particular node. You may get a scenario where you want to have more control on a node where a pod runs, such as a pod which may need high resources like heap, storage, cpu core, etc. It can be pointed to higher configuration node in a cluster,  which has dependencies that can be executed in a particular node so you may co-locate the pods based on your needs.

I had difficulties in finding the exact steps over to schedule the pod in a node. I hope it will help you. Let's see the detailed technical steps.

Create the label to run the pod in a particular slave/node using node affinity. Using "kubectl get nodes", you can get the available nodes. Choose one of the node and replace with <node-name> as shown below. Excute the below commands in the shell(Linux platform).

Command:

kubectl label nodes <node-name> <label-key>=<label-value>

Example:

kubectl label nodes ip-123-12-123-12 thisnode=slavelabel

You can check the created label using the following command.

kubectl get nodes --show-labels

Example: 

kubectl get nodes --show-labels | grep slavelabel

To delete label, 

kubectl label nodes <node-name> <label-key>-

Below JSON format works fine.  To see the result, kubectl run command with the  --overrides  option to run the pod in the particular slave node.

kubectl run -i --tty ubuntu --rm --image=ubuntu --restart=Never --overrides='
{
   "apiVersion": "v1",
   "kind": "Pod",
   "metadata": {
      "name": "ubuntutest"
   },
   "spec": {
      "affinity": {
         "nodeAffinity": {
            "requiredDuringSchedulingIgnoredDuringExecution": {
               "nodeSelectorTerms": [
                  {
                     "matchExpressions": [
                        {
                           "key": "thisnode",
                           "operator": "In",
                           "values": [
                              "slavelabel"
                           ]
                        }
                     ]
                  }
               ]
            }
         }
      }
   }
}
'

To know more about pods assignments, please refer this documentation.