Breaking down pod deployment in Kubernetes

Posted on by Simran Chawla

Kubernetes is a powerful container orchestration platform that automates the deployment, scaling, and management of containerized applications.

In this blog post, we’ll walk through the step-by-step process of deploying a pod in Kubernetes, diving into the interactions between the various components.

Process Overview:

1. Pod Creation Request:

  • The journey begins with a user or controller submitting a pod creation request using the kubectl command-line tool or the Kubernetes API.
  • The request includes the pod specification, defining the containers, resources, and other properties of the pod.

2. API Server Validation and Storage:

  • The API server receives the request and validates it against Kubernetes syntax and rules.
  • If valid, it stores the pod specification in the cluster’s data store (typically etcd).

3. Scheduler Notification:

  • The API server notifies the scheduler about the new pod, responsible for finding a suitable node in the cluster to run it.

4. Node Selection:

  • The scheduler filters available nodes based on criteria like:
    • Resource availability (CPU, memory, storage)
    • Node constraints (taints and tolerations)
    • Pod affinity and anti-affinity rules
    • Data locality

5. Node Scoring and Ranking:

  • After filtering, the scheduler scores each remaining node based on suitability for the pod.
  • Scoring criteria include factors mentioned above, plus node labels and priority classes.

6. Binding Decision:

  • The scheduler selects the node with the highest score and binds the pod to it.
  • This means updating the pod specification in the API server to include the selected node name.

7. API Server Updates and Kubelet Notification:

  • The API server propagates the binding decision to the kubelet running on the chosen node.
  • The kubelet is the agent responsible for running pods on the node.

8. Pod Creation and Execution:

  • The kubelet receives the notification and starts the process of creating the pod.
  • This involves downloading container images, creating the containers, and starting them on the node.

9. Pod Running:

  • Once the pod’s containers are up and running, the pod is marked as running in the API server, completing the deployment process.

Conclusion:

Deploying a pod in Kubernetes is a complex process involving multiple components working together. Understanding these interactions can help you appreciate how Kubernetes works and how to optimize pod deployments for performance and efficiency.