This is a multi-agent system for troubleshooting Kubernetes applications. Its initial implementation focuses on diagnosing the Open (or Advanced) Cluster Management environment, but it can be customized for other products as well.
We set up an Open Cluster Management environment to demonstrate how it works.
Install clusteradm CLI tool Run the following command to download and install the latest clusteradm command-line tool:
curl -L https://raw.githubusercontent.com/open-cluster-management-io/clusteradm/main/install.sh | bash
Setup hub and managed cluster Run the following command to quickly setup a hub cluster and 2 managed clusters by kind.
curl -L https://raw.githubusercontent.com/open-cluster-management-io/OCM/main/solutions/setup-dev-environment/local-up.sh | bash
- User: The user who ask questions and give tasks
- Executor: Execute the code written by the 'Engineer' and report the results back to them
- Engineer: Analyze the intent of the user or planner to write a sequence of shell commands or scripts
-
Planner - Kubernetes planner, responsible for making a detailed plan to accomplish a specific task within a Kubernetes environment
-
Advisor - The knowledge repository where you can find solutions and ideas for addressing any multi-cluster issues
- Manager - orchestrates the workflow between agents