Stork is Portworx's storage-aware scheduler for Kubernetes. Its architecture consists of two primary components that integrate with the native Kubernetes scheduler. The first is the Stork scheduler plugin, which runs as a separate pod and can be configured as the scheduler for specific applications. The second is a scheduler extender, which is a webhook that the default Kubernetes scheduler calls. This extender allows Stork to filter and prioritize nodes for pods using Portworx volumes, ensuring pods are co-located with their data to optimize performance and availability.

B. Stork snapshots and restores are features or operations that Stork orchestrates for data protection, not its fundamental architectural components.

C. An object store, such as an S3 bucket, is an external location used by Portworx to store backups, not an internal component of Stork.

1. Portworx Documentation

"Stork on Kubernetes": "Stork is a scheduler extender. It allows a user to co-locate pods with their data

enabling hyperconvergence. Stork also provides the ability to seamlessly migrate your applications and data from one Kubernetes cluster to another." This section explicitly identifies Stork as a scheduler extender.

2. Portworx Documentation

"Install Stork": The installation instructions detail the deployment of both the stork (controller/operator) and stork-scheduler components

confirming the two-part nature of its scheduler functionality.

3. Portworx Documentation

"Scheduler Extender": "Stork is a scheduler extender plugin for Kubernetes. When a pod is scheduled

the default Kubernetes scheduler makes an API call to the Stork extender. The Stork extender inspects the pod and if it uses Portworx volumes

it finds the best nodes for the pod to run on." This page details the function of the extender component.