Vagrant

For a more thorough introduction to Vagrant, see https://fedoraproject.org/wiki/Vagrant.

An example Vagrantfile is provided as Vagrantfile.example. To use it, just copy it and install Vagrant:

$ cp dev/Vagrantfile.example Vagrantfile
$ sudo dnf install ansible libvirt vagrant-libvirt vagrant-sshfs vagrant-hostmanager
$ vagrant up

If you get this error Block in synced_folders: Internal error. Invalid: sshfs, when you run vagrant up , you need to install vagrant sshfs plugin, which can be done by:

$ vagrant plugin install vagrant--sshfs

and then:

$ vagrant up

The default Vagrantfile forwards ports from the host to the guest, so you can interact with the application as if it were running on your host machine.

When the vagrant VM is up and running, connect to it with:

$ vagrant ssh

This will log you into the VM as the user vagrant which has a couple of aliases preconfigured:

$ pstart            # Starts pagure, the workers and other tasks
$ pstop             # Stops all those tasks again
$ pstatus           # Shows pagure status

The Vagrant pagure doesn't have its own log file, use journalctl -f to show the pagure output. The verbosity can be configured in the pagure config file with the LOGGING parameter.

Docker Compose

Create the folder that will receive the projects, forks, docs, requests and tickets' git repo:

mkdir -p lcl/{repos,remotes,attachments,releases}

A docker compose environment is available to run pagure. First use the following command to build the containers.

$ docker-compose -f dev/docker-compose.yml build

Once all the containers are built, run the following command to start the containers.

$ docker-compose -f dev/docker-compose.yml up -d

Once all the containers have started, you can access pagure on http://localhost:5000

To stop the containers, run the following

$ docker-compose -f dev/docker-compose.yml stop

More information about docker-compose cli see https://docs.docker.com/compose/reference/.

$ ./dev/run-tests-docker.py

$ ./dev/run-tests-docker.py --centos

$ ./dev/run-tests-docker.py --skip-build
$ ./dev/run-tests-docker.py --centos --skip-build

$ ./dev/run-tests-docker.py tests/test_pagure_flask_ui_priorities.py

$ ./dev/run-tests-docker.py tests/test_pagure_flask_ui_priorities.py:PagureFlaskPrioritiestests.test_ticket_with_no_priority

$ ./dev/run-tests-docker.py --help

Manually

• Install the needed system libraries:

  sudo dnf install git python3 python3-devel libgit2-devel redis \
                   libjpeg-devel gcc libffi-devel redhat-rpm-config
  

  Note

  Do note the version of libgit2 that you install, for example in libgit2-0.23.4-1 you need to keep in mind the 0.23

  Note

  On RHEL and derivative (CentOS, Scientific Linux) there is no python3 package. Just python36 or python34 available in EPEL 7 (EPEL 6 only has python34). Choose the one you prefer (3.6 is newer and generally a better choice).

• Retrieve the sources:

  git clone https://pagure.io/pagure.git
  cd pagure
  

• Install dependencies

  • create the virtual environment (use pytohn3.X explicitly on EPEL):

    python3 -m venv pagure_env
    source ./pagure_env/bin/activate
    

  • Install the correct version of pygit2:

    pip install pygit2==<version of libgit2 found>.*
    

    So in our example:

    pip install pygit2==0.23.*
    

  • Install the rest of the dependencies:

    pip install -r requirements.txt
    

• Create the folder that will receive the projects, forks, docs, requests and tickets' git repo:

  mkdir -p lcl/{repos,remotes,attachments,releases}
  

• Copy and edit the alembic.ini file (especially the script_location key):

  cp files/alembic.ini .
  vim alembic.ini
  

• Set the script_location to alembic, ie: the folder where the revisions are stored, relative to the location of the alembic.ini file.

• Create the inital database scheme:

  python createdb.py --initial alembic.ini
  

• Enable and start redis server:

  sudo systemctl enable redis
  sudo systemctl start redis
  

• Start a worker, in one terminal:

  ./runworker.py
  

• Run the application, in another terminal:

  ./runserver.py
  

• To get some profiling information you can also run it as:

  ./runserver.py --profile
  

This will launch the application at http://127.0.0.1:5000

• To run unit-tests on pagure

  • Install the dependencies:

    pip install -r tests_requirements.txt
    

  • Run it:

    python runtests.py run
    

  You may use:

  python runtests.py --help
  

  to check other options supported or read the source code ;-)

  Note

  While testing for worker tasks, pagure uses celery in /usr/bin/ Celery then looks for eventlet (which we use for testing only) at system level and not in virtual environment. You will need to install eventlet outside of your virtual environment if you are using one.