AutoTest

Runs Docker container instances that grade student code for particular classes. Leaves Github comments online with Docker container result information when requested in a comment such as @autobot #d1 by a student on a comment underneath a commit on Github online.

Requirements

• MongoDB db version v3.4.7 (any close version)
• Node JS v6.11.0 LTS (higher will likely work)
• Yarn 0.24.6 (any close version)
• Docker version 17.03.1-ce, build c6d412e (any close version)
• 'autotest.env' file in ./autotest/ directory:

	GITHUB_API_KEY=LONGSTRING
	REDIS_ADDRESS=http://localhost
	SSL_CERT=/path/to/crt.crt
	SSL_KEY=/path/to/key.key
	SSL_INT_CERT=/path/to/-cacerts.pem
	COURSES=210 310 etc.
	DEV_MONGO_DB_INSTANCE=mongodb://username:password@localhost:27017/development
	PROD_MONGO_DB_INSTANCE=mongodb://username:password@localhost:27017/production
	TEST_MONGO_DB_INSTANCE=mongodb://username:password@localhost:27017/testing
	MENTION_TAG=@autobot
	

• In dir path, yarn run install with original lockfile and then yarn run build
• GITHUB_API_KEY may be created on Github under 'Settings' with permissions to manage orgs, repos, repo_hooks, and users.

Start Instructions

• ./deploy.sh Install required Docker instances (must be run after COURSES env set)
• ./autotest/docker/tester/docker-up.sh Build required per-course Docker containers.
• yarn run start Start AutoTest service.

Extra Info

• Based on the courses in the autotest.env file, an instance of the application will start on a corresponding port that begins in the 1000s to get around sudo security requirements under port 1000.