Node selectors for simple scheduling in Kubernetes - Step-by-Step Execution

Process Flow - Node selectors for simple scheduling

Pod Spec Created

↓

Add nodeSelector Field

↓

Pod Submitted to API Server

↓

Scheduler Checks Nodes

↓

Match Node Labels with nodeSelector?

No→Wait or Fail

Yes↓

Pod Assigned to Matching Node

↓

Pod Runs on Selected Node

The pod spec includes a nodeSelector field. The scheduler matches this with node labels to assign the pod to a suitable node.

Execution Sample

Kubernetes

apiVersion: v1
kind: Pod
metadata:
  name: example-pod
spec:
  nodeSelector:
    disktype: ssd
  containers:
  - name: app
    image: nginx

This pod requests scheduling only on nodes labeled with disktype=ssd.

Process Table

Step	Action	Scheduler Check	Result	Pod State
1	Pod spec created with nodeSelector disktype=ssd	N/A	Pod ready for scheduling	Pending
2	Pod submitted to API server	N/A	Pod accepted	Pending
3	Scheduler scans nodes	Check node labels	Found nodes with disktype=ssd	Pending
4	Scheduler matches nodeSelector with node labels	Match disktype=ssd?	Match found on node-1	Pending
5	Scheduler assigns pod to node-1	Assignment done	Pod bound to node-1	Running
6	Pod starts running on node-1	N/A	Pod running	Running

💡 Pod assigned to node with matching label disktype=ssd and starts running.

Status Tracker

Variable	Start	After Step 3	After Step 5	Final
Pod State	Not created	Pending	Running	Running
Node Selected	None	None	node-1	node-1

Key Moments - 3 Insights

Why does the pod stay in Pending state before scheduling?

The pod is Pending because the scheduler has not yet assigned it to a node. See execution_table step 2 and 3 where the pod is accepted but not assigned.

What happens if no node matches the nodeSelector labels?

The pod remains Pending and will not be scheduled until a matching node appears. This is shown by the scheduler check in step 4 where no match would cause waiting.

Can the pod run on a node without the matching label?

No, the scheduler enforces nodeSelector strictly. The pod only runs on nodes matching the labels, as shown in step 5 and 6.

Visual Quiz - 3 Questions

Look at the execution table, at which step does the scheduler assign the pod to a node?

A Step 4

B Step 3

C Step 5

D Step 6

💡 Hint

Check the 'Result' column for the assignment action in the execution_table.

According to the variable tracker, what is the pod state after step 3?

A Pending

B Running

C Succeeded

D Failed

💡 Hint

Look at the 'Pod State' row and the 'After Step 3' column in variable_tracker.

If the nodeSelector label was changed to disktype=hdd, what would happen in the execution table?

A Scheduler finds matching nodes and assigns pod immediately

B Pod remains Pending if no node has disktype=hdd label

C Pod runs on any node regardless of label

D Pod is deleted automatically

💡 Hint

Refer to key_moments about what happens if no node matches the nodeSelector.

Concept Snapshot

Node selectors let you tell Kubernetes to run pods only on nodes with specific labels.
Add a nodeSelector field in pod spec with key-value pairs.
Scheduler matches these labels to node labels.
Pod stays Pending if no matching node is found.
Simple way to control pod placement without complex rules.

Full Transcript

Node selectors are a simple way to schedule pods on specific nodes in Kubernetes. You add a nodeSelector field to the pod specification with key-value pairs that must match node labels. When the pod is submitted, the scheduler looks for nodes with matching labels. If it finds one, it assigns the pod to that node and the pod starts running. If no matching node exists, the pod stays in Pending state until a suitable node appears. This method helps control where pods run using simple label matching.