Hi! We're excited to have you as a contributor.
Have questions about this document or anything not covered here?
Come chat with us at #ansible-community on irc.freenode.net.
- All code submissions are done through pull requests
against the
develbranch. - You must use
git commit --signofffor any commit to be merged, and agree that usage of--signoffconstitutes agreement with the terms of DCO 1.1. - Take care to make sure no merge commits are in the submission, and use git rebase vs git merge for this reason.
- If submitting a large code change, it's a good idea to join the
#ansible-communitychannel onirc.freenode.net, and talk about what you would like to do or add first. This not only helps everyone know what's going on, it also helps save time and effort, if the community decides some changes are needed. - We ask all of our community members and contributors to adhere to the Ansible code of conduct. If you have questions, or need assistance, please reach out to our community team at [email protected].
The Galaxy development environment workflow and toolchain is based on Docker,
and the docker-compose tool, to provide dependencies, services,
and databases necessary to run all of the components.
It also binds the local source tree into the development container,
making it possible to observe and test changes in real time.
For convenience, many of the commands you'll be running to build,
start, stop and interact with the development containers have been added
to the project Makefile.
Unless you really like typing, you'll want to take advantage of this.
Check that you have access to the make command, and if not,
install the make package for your OS.
Prior to starting the development services, you'll need docker
and docker-compose. On Linux, you can generally find these in your
distro's packaging, but you may find that Docker themselves maintain
a separate repo that tracks more closely to the latest releases.
For MacOS and Windows, we recommend Docker for Mac and Docker for Windows respectively.
For Linux platforms, refer to the following from Docker:
Fedora
https://docs.docker.com/engine/installation/linux/docker-ce/fedora/
Centos
https://docs.docker.com/engine/installation/linux/docker-ce/centos/
Ubuntu
https://docs.docker.com/engine/installation/linux/docker-ce/ubuntu/
Debian
https://docs.docker.com/engine/installation/linux/docker-ce/debian/
Arch
https://wiki.archlinux.org/index.php/Docker
If you have not done so already, you'll need to fork the Galaxy repo on GitHub. For more on how to do this, see Fork a Repo.
The first step to running a local Galaxy instance is building the images.
You're going to run a script that will build the galaxy-dev image,
which will contain everything needed to run the Galaxy frontend web server,
backend Django server, and Celery task runner.
Prior to building this image, the script will first build the galaxy-build
image, which contains all of the required OS packages, Python dependencies,
and frontend tools and packages. The galaxy-build image is then used
as the base image to create galaxy-dev.
If you're curious about what actually goes into building the images,
you'll find the Dockerfile for galaxy-build in
scripts/docker/release.
The actual filename is Dockerfile.build. And you'll find the Dockerfile
for galaxy-dev in scripts/docker/dev.
Run the following to build the image:
## Set your working directory to the project root
$ cd galaxy
## Start the build process
$ make dev/buildOnce the build completes, you will have the galaxy-dev and galaxy-build
images in your local image cache. Use the docker images command to check,
as follows:
$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
galaxy-dev latest 2f552729e204 22 seconds ago 748MB
galaxy-build latest c456f5c226d3 2 minutes ago 568MB
centos 7 196e0ce0c9fb 6 weeks ago 197MBWe use docker-compose to run the containers. If you're curious about
the services, and and how they're configured, the compose file is
scripts/docker/dev/compose.yml.
There are a couple different ways to start the development containers.
If all of this is new, and you just want to get things going, run the
following command to start the containers in an attached mode.
All that means is that the stdout and stderr for each container will
stream to the stdout and stderr of your terminal session.
After running the command, your session will be totally consumed with the output, which is OK. Seeing the output lets you know what's actually happening in the containers. So afterwards, to run additional commands from your terminal, you'll need to start a second session.
So without further ado, run the following:
## Set your working directory to the project root
$ cd galaxy
## Start the build process
$ make dev/upAny missing images (i.e. postgresql, rabbitmq, prometheus, influxdb, grafana) will be pulled. Getting all the images downloaded may take a few minutes. Once all the images are available, the ontainers will launch.
Aftr the above commands completes, you can take a look at the containers by
running docker ps in your second terminal session:
$ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
b76488f94890 galaxy-dev:latest "/entrypoint.sh /g..." 2 minutes ago Up 2 minutes 0.0.0.0:8000->8000/tcp galaxy_galaxy_1
dfe97d19197e centos/postgresql-95-centos7 "container-entrypo..." 22 hours ago Up 2 minutes 0.0.0.0:2345->5432/tcp galaxy_postgres_1
fd3dd5f663f2 rabbitmq:latest "docker-entrypoint..." 22 hours ago Up 2 minutes 4369/tcp, 5671-5672/tcp, 25672/tcp galaxy_rabbitmq_1
9561d0cea1ec prom/prometheus:latest "/bin/prometheus -..." 2 minutes ago Up 2 minutes 0.0.0.0:9090->9090/tcp galaxy_prometheus_1
21e8b688f2ab influxdb:latest "/entrypoint.sh in..." 22 hours ago Up 2 minutes 0.0.0.0:8086->8086/tcp galaxy_influxdb_1
92186c792b4d grafana/grafana:latest "/run.sh" 2 minutes ago Up 2 minutes 0.0.0.0:3000->3000/tcp galaxy_grafana_1If you prefer to start the containers in detached mode, where they run in the background, run the following command:
## Set your working directory to the project root
$ cd galaxy
## Start the build process
$ make dev/up_detachedSince the stdout and stderr are not streaming to your terminal
session, you'll need to use the docker logs command to view logging output.
As pictured above in Quick start, use docker ps to see
the list of running containers, then use docker logs -f <container name>
to stream a container's output. Use <Ctrl-C> to stop the streaming output.
If you're familiar with tmux, and you would prefer to view the container
output through a tmux session, use the following 2-step process to launch
the containers, and then start the processes within the galaxy service
using tmux.
Execute the following to launch the containers in detached mode, running in the background. Note that there will be no running processes in the
galaxyservice container, because you'll start them in the next step withtmux.## Set your working directory to the project root $ cd galaxy ## Start the build process $ make dev/up_tmux
Once the above commands completes, you can view the containers by running
docker ps. The service we're most interested in isgalaxy, and it's container name will begalaxy_galaxy_1. Before we can launchtmux, we need to wait for database migrations and other setup to complete. To see what's going on inside thegalaxyservice container, and whether or not the setup is complete, run the following to stream its logging output:$ docker logs -f galaxy_galaxy_1The above will stream the log output to your terminal window. When all the migrations and setup work is done, the output stream will stop, and you'll see output similar to the following:
Applying socialaccount.0003_extra_data_default_dict... OK + make build_indexes Rebuild Custom Indexes /var/lib/galaxy/venv/bin/python ./manage.py rebuild_galaxy_indexes /var/lib/galaxy/venv/lib/python2.7/site-packages/maintenance/middleware.py:3: RemovedInDjango19Warning: django.utils.importlib will be removed in Django 1.9. from django.utils.importlib import import_module Rebuild Search Index /var/lib/galaxy/venv/bin/python ./manage.py rebuild_index --noinput /var/lib/galaxy/venv/lib/python2.7/site-packages/maintenance/middleware.py:3: RemovedInDjango19Warning: django.utils.importlib will be removed in Django 1.9. from django.utils.importlib import import_module Removing all documents from your index because you said so. All documents removed. Indexing 0 roles + '[' 1 == 1 ']' + scripts/docker/dev/sleep.sh
Once you see the very last line,
scripts/docker/dev/sleep.sh, you're ready for the next step.Now you'll start
tmuxand launch the processes inside thegalaxyservice container by running the following. If you're streaming the logging ouptut still, use<Ctrl-C>to stop the stream.## Set your working directory to the project root $ cd galaxy ## Start tmux $ make dev/tmux
After doing all this work, you're anxious to view your shiny new Galaxy site,
aren't you? Well, not so fast. As discussed above in Running through tmux,
you'll need to first check to make sure that all the database migrations
and setup work completed, and processes are running inside the galaxy
service container.
Check the output stream from the galaxy service container, and look for
the completion of database migrations, and the start of the gulp web server.
If you see output similar to the following, then you know that gulp is
running and accepting connections:
[03:10:00] Using gulpfile /galaxy/gulpfile.js
[03:10:01] Starting 'less'...
[03:10:01] Starting 'server'...
[HPM] Proxy created: / -> http://localhost:8888
[03:10:01] Finished 'server' after 130 ms
[03:10:01] Starting 'watch'...
[03:10:03] Finished 'watch' after 2.06 s
[Browsersync] Access URLs:
-----------------------------------
Local: http://localhost:8000
External: http://172.18.0.6:8000
-----------------------------------
UI: http://localhost:3001
UI External: http://172.18.0.6:3001
-----------------------------------
[Browsersync] Serving files from: /galaxy
[03:10:04] Finished 'less' after 3.23 s
[03:10:04] Starting 'default'...
[03:10:04] Finished 'default' after 108 μs
OK, go for it! Your Galaxy web site is available at: http://localhost:8000.
Note
You won't be able to authenticate until you perform the post build steps for creating an admin user and configuring GitHub authentication.
From the root of the project tree, run make dev/setup-metrics to initialize
InfluxDB and import data sources and dashboards for Prometheus and InfluxDB into Grafana.
$ make dev/setup-metricsLog into Grafana at http://localhost:3000 using the admin user with
password admin. Navigate to Galaxy Search Metrics - InfluxDB and
Galaxy Search Metrics - Prometheus dashboards.
Search metrics are exposed at http://localhost:8000/metrics. From there, the metrics are being scraped by Prometheus (http://localhost:9090/). Prometheus serves as a short-term storage. For a long-term metrics storage, the metrics are being sent from Prometheus to InfluxDB.
From the root of the project tree, run make dev/createsuperuser to start
the creation process. You'll be prompted for the vital details as depicted below:
$ make dev/createsuperuser
Create Superuser
Username: admin
Email address: [email protected]
Password:
Password (again):
Superuser created successfully.The Django admin site can be accessed at http://localhost:8000/admin.
To log into the development site, you first have to authorize it as a
GitHub Oauth Application. You can do this by logging into GitHub,
going to Personal Settings, choosing OAuth Applications,
and then doing the following to create a new app:
- Click Register New Application.
- Set the Homepage URL to
http://localhost:8000. - Set the Authorization Callback URL to
http://localhost:8000/accounts/github/login/callback/.
After you save the new application, access your local Galaxy admin site at http://localhost:8000/admin, and log in using the admin user you created above in [Create admin user](#create-admin-user)
Click on Sites. You'll see one site defined, example.com.
Click on example.com to modify it. On the next page, change both the
Domain Name and Display Name from example.com to localhost.
Click the Save button.
Next, create a new social application. Start by finding Social applications
at the bottom of the list, on the admin site home page.
Click the Add button to its right, and on the next page, complete the
following to configure the new application:
- Set the Provider to
GitHub. - Enter
GitHubas the Name. - From the new GitHub OAuth application you just created, copy the ClientID value into Client id.
- Copy the Client Secret value into Secret key.
- Under Sites, add
localhostto Chosen sites. Save the changes.
Now test the authentication. Log out of your admin account, and go back to the home page at http://localhost:8000. Now log in using your GitHub account by clicking the GitHub logo under Log into Galaxy with GitHub.
The Javascript, CSS and HTML components for the web site can be found in the
galaxy/static folder. Within this folder,
the gulp service watches for modifications to less/*.less stylesheets,
and automatically recompiles the CSS and refreshes your browser.
It also refreshes your browser whenever changes are made
to js/*/*.js and partion/*.html files.
To stop all services, run make dev/down.
Once you have galaxy composed and running, you may also run different commands to check your changes.
To do this you need galaxy running in detached state or run commands from separate terminal session.
Full list of commands is available in Makefile, however we want to highlight the most useful here.
To run lint checks against Python sources, execute:
$ make dev/flake8To run lint checks against JavaScript/TypeScript sources, execute:
$ make dev/jslintWe use prettier to enforce code formatting for all of our TypeScript and less files. To automatically format your Angular code run:
$ make dev/prettierUnformatted code will cause the travis build to fail when you push your changes to GitHub.
It's recommended that you set up prettier on your editor if you're making lots of
changes to anything in galaxyui/. Prettier is supported by most major editors
and you can find more information about that here.
Our prettier configuration can be found at galaxyui/.prettierrc.yaml. Please
use it when setting up your editor.
To run unit and functional tests against execute:
$ make dev/testThis command will test Python code and also will produce test coverage reports.
There are 3 kinds of reports produced:
- Console report, that shows coverage in console
- Static HTML files located in htmlcov with htmlcov/index.html as entry point
coverage.xml, that may be used by your IDE or text editor.
These files are not part of git repository and will not be commited.
If you use VSCode as your editor and want to integrate coverage.xml
report and highlight code accordingly, then you may be interested in
Coverage Gutters
plugin. In order to make it working, you need to install it and add following lines into your VSCode configuration:
"coverage-gutters.xmlname": "coverage.xml",
If you use PyCharm Professional, then you may configure it according to the Code Coverage guide.
Static HTML files may be used simply by opening them in your favourite web browser.
Once you have all lint and tests passed and you are ready to commit your patch and propose pull request, install pre-commit hook, provided by this repository:
$ cp pre-commit .git/hooks/pre-commitDon't forget to commit your code with git commit --signoff as described in
the top of this document and follow other guidelines.
Thank you for your contribution!