815_krknPR_about_disk_failure_simulation

[10sharmashivam]

Description

This PR adds disk failure simulation functionality to krkn-lib, which will be used by the main krkn repository for disk failure chaos scenarios. The implementation uses Kubernetes-native approaches instead of direct oc debug commands.

Context

This PR addresses the following comment from the main krkn repository: PR

"We try to keep krkn as Kubernetes-based as possible. Can you try using this function in krkn-lib instead of oc debug to be sure this functionality can work on all cluster types?"

The original implementation used OpenShift-specific oc debug commands, which limited the functionality to OpenShift clusters only. This PR moves the functionality to krkn-lib and implements it using pure Kubernetes-native approaches.

Why a new PR in krkn-lib?

1. Architectural Decision: krkn-lib is designed to contain shared functionality that works across different Kubernetes distributions. Moving the disk failure simulation here follows this architectural pattern.

2. Distribution Agnostic: The current implementation in krkn uses OpenShift-specific commands (oc debug). Moving this to krkn-lib and using Kubernetes-native approaches makes it work across all Kubernetes distributions.

3. Code Reusability: This functionality can be reused by other components or projects that use krkn-lib.

Changes

• Added simulate_disk_failure method to KrknKubernetes class
• Replaced OpenShift-specific oc debug commands with Kubernetes-native approaches:
  • Uses privileged pods for node operations
  • Uses sysfs for disk operations
  • Uses standard Kubernetes volume mounts
• Includes proper validation and error handling
• Follows Kubernetes best practices for pod management
• Maintains security with read-only mounts and proper cleanup

Related Issues and PR

PR 815
Issues 1 & 2