Tango is a standalone RESTful Web service that runs and manages jobs. A job is a set of files that must satisfy the following constraints:

1. There must be exactly one Makefile that runs the job.
2. The output for the job should be printed to stdout.

Example jobs are provided for the user to peruse in clients/. Tango has a REST API which is used for job submission.

Upon receiving a job, Tango will copy all of the job's input files into a VM, run make, and copy the resulting output back to the host machine. Tango jobs are run in pre-configured VMs. Support for various Virtual Machine Management Systems (VMMSs) like KVM, Docker, or Amazon EC2 can be added by implementing a high level VMMS API that Tango provides.

A brief overview of the Tango respository:

• tango.py - Main tango server
• jobQueue.py - Manages the job queue
• jobManager.py - Assigns jobs to free VMs
• worker.py - Shepherds a job through its execution
• preallocator.py - Manages pools of VMs
• vmms/ - VMMS library implementations
• restful_tango/ - HTTP server layer on the main Tango

Tango was developed as a distributed grading system for Autolab at Carnegie Mellon University and has been extensively used for autograding programming assignments in CMU courses.

Please feel free to use Tango at your school/organization. If you run into any problems with the steps below, you can reach the core developers at [email protected] and we would be happy to help.

1. Follow the steps to set up Tango.
2. Read the documentation for the REST API.
3. Read the documentation for the VMMS API.
4. Test whether Tango is set up properly and can process jobs.

Tango now runs on Python 3. However, there is a legacy branch master-python2 which is a snapshot of the last Python 2 Tango commit for legacy reasons. You are strongly encouraged to upgrade to the current Python 3 version of Tango if you are still on the Python 2 version, as future enhancements and bug fixes will be focused on the current master.

We will not be backporting new features from master to master-python2.

1. Fork the Tango repository.
2. Create a local clone of the forked repo.
3. Install pre-commit from pip, and run pre-commit install to set up Git pre-commit linting scripts.
4. Make a branch for your feature and start committing changes.
5. Create a pull request (PR).
6. Address any comments by updating the PR and wait for it to be accepted.
7. Once your PR is accepted, a reviewer will ask you to squash the commits on your branch into one well-worded commit.
8. Squash your commits into one and push to your branch on your forked repo.
9. A reviewer will fetch from your repo, rebase your commit, and push to Tango.

Please see the git linear development guide for a more in-depth explanation of the version control model that we use.

Tango is released under the Apache License 2.0.