opsimulate

Georgia Tech OMSCS EduTech Class Project

Description

The opsimulate CLI allows students to deploy a working, accessible Gitlab instance and then start a simulation of a production incident with that Gitlab instance that they can investigate, debug, and fix. The opsimulate CLI will have the functionality to create servers, setup the server packages and configuration in a specific way, setup and run a Gitlab service, provide the student with SSH access to the server, and setup an incident from an incident module. The opsimulate CLI tool will also provide hints and a description of and also solution for (if asked for) the incident. It can also check whether the student has successfully resolved the incident.

Motivation

There are currently no free, open-source simulation-based learning frameworks or tools available for gaining sysadmin/devops related technical operational skills, knowledge, and experience.

Dependencies

• Python 2.x
• a Google Compute Platform Account

Installation

pip install opsimulate

Google Compute Platform Setup

• Sign into your GCP account using a Google account at console.cloud.google.com (Gmail works). Please create a Google account if you do not have one.

• At the GCP Home Dashboard, create a project.

• Go the IAM --> Service Accounts page.

• Create a service account. Make sure you select the option to "Furnish a new private key" of the "JSON" Key Type.

Also, make sure you give the service account the Project Owner role.

• Creating the service account will automatically download the service account credentials. It will download as a json file. Note where the credentials json file downloaded to because it will be required when you run opsimulate load_credentials to properly copy over the GCP credentials.

Usage/Workflow

• Run opsimulate setup to setup the home ~/.opsimulate directory that opsimulate generates artifacts in.
• Run opsimulate load_credentials <path to GCP-service-account-credentials.json> to load in the GCP service account credentials.
• Run opsimulate deploy to create the Gitlab instance and make it accessible.
• Run opsimulate status until the output displays that both the VM and Gitlab are up and ready.
• Select a problem module via opsimulate module_select <path to problem module dir>
• Start the problem simulation via opsimulate module_start
• In a separate command line window, run opsimulate connect and run the printed SSH command to connect to your Gitlab server.
• Investigate, debug, fix, do whatever you need to figure out what the issue is and how to fix it.
• If you need help/hints, run opsimulate module_hint as many times as you need.
• If at any time, you want to check if the problem has been fixed, run opsimulate module_check.
• If you can't fix the issue yourself, run opsimulate module_resolve to fix the issue and end the problem simulation. Your Gitlab instance will still be up.
• When you're completely done, run opsimulate clean to clean up local artifacts (ex. generated SSH keys) and tear down GCP resources like the Gitlab VM.

Commands

• opsimulate status = Check the status of GCP credentials, selected problem module, VM and Gitlab status
• opsimulate setup = Setup the ~/.opsimulate artifacts directory
• opsimulate load_credentials = Copy over the GCP service account credentials into ~/.opsimulate directory
• opsimulate deploy = Deploy the GCP VM with Gitlab and setup SSH access
• opsimulate connect = Print the SSH command that can be used to get onto the Gitlab VM
• opsimulate clean = Clean up the GCP Gitlab VM and generated SSH keys
• opsimulate module_select = Select a problem module
• opsimulate module_start = Start the problem simulation
• opsimulate module_hint = Get a hint about the problem
• opsimulate module_check = Check if the problem has been resolved
• opsimulate module_resolve = Resolve the problem

Contributing

See the CONTRIBUTING.md

Docker Workflow

Running/testing opsimulate source:

docker build -t opsimulate-testing -f Dockerfile.testing .
docker run -v <absolute path to service account json>:/root/.opsimulate/service-account.json -v <absolute path to module>:/tmp/module -it opsimulate-testing bash

Running opsimulate (from Pypi) in a container:

docker build -t opsimulate .
docker run -v <absolute path to service account json>:/root/.opsimulate/service-account.json -v <absolute path to module>:/tmp/module -it opsimulate bash

Pypi Build/Upload Workflow

python setup.py sdist
python setup.py sdist upload -r pypitest
python setup.py sdist upload -r pypi