Skip to main content

Chaos faults for SSH

Last updated on

Introduction

SSH chaos injects user-defined chaos on a remote VM by executing custom scripts (inject and rollback) over SSH. Use it when the chaos you want to run is not covered by a first-class fault, when you need to run a pre-built script that already exists in your operations toolkit, or when you want a portable framework for the long tail of one-off resilience tests.

SSH chaos runs a user-supplied inject script on a remote host over SSH for a configurable duration and runs a rollback script when the duration ends or the experiment is aborted. Scripts are passed through a ConfigMap; SSH credentials (SSH_PASSWORD and SSH_KEY) are referenced from Harness Secret Manager.

Details

Use cases Use as a customizable framework when the long-tail chaos you need (custom network rules, custom service restarts, custom database failovers) is not covered by a first-class fault. The rollback script ensures the target host returns to baseline on completion or abort.

View details