Skip to content

Latest commit

 

History

History
160 lines (135 loc) · 6.43 KB

GUIDE.md

File metadata and controls

160 lines (135 loc) · 6.43 KB
Raw

TAG - New Joiner onboarding - 2022 edition

A few useful tips for TAG new joiners. SIO2Project is a growing project mainly maintained by the TAG team. The project's architecture was previously held on self-served platforms such as Gerrit, Hudson etc. By the end of academic year 2022 most of the architecture was moved to GitHub.

  • Gerrit (code review) -> GitHub code review
  • Hudson (automated jobs) -> GitHub Actions
  • Jira (ticketing system) -> GitHub Issues

There are still some deployment jobs left on Hudson, but GitHub Secrets can be a bright future for OIOIOI. All the tickets raised during development are still in Jira, and it is not intended to change that.

Prepare your development environment

First things first - the best way to run OIOIOI (web service of SIO2Project) is to use Docker. Any other installation method described in documentation does not work, so you just need to follow instructions in README.rst. Using docker-compose works only on Linux and MacOS - Windows is not supported.

Install Docker and docker-compose on your computer - if you don't know how just type in How to install Docker on XZY, where XYZ is your operating system. Do the same for docker-compose.

Remember to add user to docker group

sudo groupadd docker
gpasswd -a $USER docker
newgrp docker

All the commands below are being run in oioioi directory (main directory of the repository). In order to use easy_toolbox.py alternative method, check python package requirements in easy_toolbox.py.

To build OIOIOI image run

OIOIOI_UID=$(id -u) docker-compose -f docker-compose-dev.yml build

or

./easy_toolbox.py build

Set your containers up and running

OIOIOI_UID=$(id -u) docker-compose -f docker-compose-dev.yml up -d

or

./easy_toolbox.py up

Wait some time for the migration to finish (no more than a few minutes).

Run your web service

OIOIOI_UID=$(id -u) docker-compose -f docker-compose-dev.yml exec web python3 manage.py runserver 0.0.0.0:8000

or

./easy_toolbox.py run

Now visit localhost:8000 and start exploring OIOIOI.

Run unit tests

In order to run unit tests Docker installation is required. To do it just run

docker-compose -f docker-compose-dev.yml exec "web" ../oioioi/test.sh

or

./easy_toolbox.py test

Run e2e tests

In order to run Cypress tests a few more steps are required.

  • Clear the database
  • Create superuser (admin, admin)
  • Run the service
  • Run the tests

All of these steps can be done by running

./easy_toolbox.py flush-db && ./easy_toolbox.py add-superuser && ./easy_toolbox.py cypress-apply-settings && ./easy_toolbox.py run

and (in new terminal)

./test_cypress.sh [-g]

Normally these tests are being run on GitHub Actions as a nightly test workflow, but in the future it is intended to create separate testing environment for Cypress using SQLite database and data from fixtures files.

Contributing

In order to contribute to OIOIOI just fork the repository, clone the copy, make changes as in normal repository and when the change is ready, create a push request to master branch of sio2project/oioioi.

GitHub Actions

As mentioned above GitHub Actions replaced most of the Hudson jobs. Right now the available workflows are:

  • Unit tests
  • Nightly tests (unit tests with testes marked as 'slow')
  • Bi-Nightly Cypress e2e tests
  • Documentation generator
  • Transifex translations downloader
  • Transifex translations uploader

Deployment

No changes can be seen on szkopul.edu.pl without deployment. @twalen is responsible for conducting the deployments, and they are usually done during maintenance window (see szkopul.edu.pl main page for details).

During deployment translation files are being uploaded and downloaded. The translation manager is Transifex. When you add localized text to OIOIOI it is uploaded to Transifex with GitHub Action and later, when translated, downloaded, compiled and saved to the codebase. (Both of these jobs need to be run manually).

Ticketing

All the ticketing is still done via Jira. Ticketing was moved to GitHub Issues.

FAQ

When to use ./easy_toolbox.py build?

Essentially, it is not easy to answer. Firstly, let's understand what each of the commands does. As you may already know, SIO2 development environment works in containerized infrastructure. We have four containers up and running, required for SIO2 to work properly. Database (db), RabbitMQ (broker), Worker (worker) and OIOIOI (web). Postgres and RabbitMQ already have existing docker images on Docker Hub. We only need to build (here comes the magic word) SIO dependent images. Dockerfile defines what steps are to be done in order to create the environment. Once you built the image, you can set the container up. Remember - things like dependencies (requirements[_static].txt, setup.py) are downloaded during the built, so if you changed something in those places you either need to build the image again, or apply these changes by hand (e.g. do pip install). If you have good internet connection and adequate CPU, it shouldn't be hard to build the image again, especially that it is more stable approach.

When to use ./easy_toolbox.py up?

Whenever you don't have the containers up. After running it you should see the following output:

Running command OIOIOI_UID=$(id -u) docker-compose -f docker-compose-dev.yml up -d
===================================================================================================================================
Creating network "oioioi_default" with the default driver
Creating oioioi_broker_1 ... done
Creating oioioi_db_1     ... done
Creating oioioi_web_1    ... done
Creating oioioi_worker_1 ... done

If not, check the output of docker ps -a and docker logs <container_name>.

When to use ./easy_toolbox.py run?

Whenever you want the Django web service to be running. In development environment we use dev server, so it should catch all changes in the code. You can have the server running all the time - just make sure, that Django discovered your changes (the server will restart with appropriate message like "detected changes in file xyz.py").

What to do when I get permission denied error?

Check if you added user to docker group. For more information check 'Prepare your development environment' section.