tutor-discovery 18.0.0

A Tutor plugin for course discovery, the Open edX service for providing access to consolidated course and program metadata

This is a plugin for Tutor that integrates the Course Discovery application in an Open edX platform. it is useful for integration with, for example, Ecommerce or an external course catalog.

Installation

This plugin requires tutor>=3.6.0. If you have installed tutor from a pre-compiled binary, it already comes bundled with the discovery plugin. Otherwise:

tutor plugins install discovery

Then, to enable this plugin, run:

tutor plugins enable discovery

Restart and initialize your platform with:

tutor local launch

Operations

Creating a user

The discovery user interface will be available at http://discovery.local.edly.io for a local test instance, and at DISCOVERY_HOST (by default: http(s)://discovery.<your lms host>) in production. In order to run commands from the UI, a user must be created:

tutor local run discovery ./manage.py createsuperuser

Then, you must login with this user at http://discovery.local.edly.io/admin.

Alternatively, you can login with oauth2 using a pre-existing user created on the LMS/CMS by accessing http(s)://discovery.<your lms host>/login. To do so, the proper domain names must exist and point to the production server.

Index configuration

Discovery uses separate indices for different models (the names are: course, course_run, person and program by default). And you can overwrite theses names by configuring DISCOVERY_INDEX_OVERRIDES:

DISCOVERY_INDEX_OVERRIDES:
  course_discovery.apps.course_metadata.search_indexes.documents.course: your-course-index-name
  course_discovery.apps.course_metadata.search_indexes.documents.course_run: your-course-run-index-name
  course_discovery.apps.course_metadata.search_indexes.documents.person: your-person-index-name
  course_discovery.apps.course_metadata.search_indexes.documents.program: your-program-index-name

Re-indexing courses

tutor local run discovery ./manage.py refresh_course_metadata --partner_code=openedx
tutor local run discovery ./manage.py update_index --disable-change-limit

Caching programs

In order to cache programs in the LMS, you will need to manually create a catalog integration. Make sure you use staff user for the below command. If lms_catalog_service_user is not a staff user, then make it a staff user in your LMS User model. This step should be performed just once:

tutor local run lms ./manage.py lms create_catalog_integrations --enabled \
    --internal_api_url="" \
    --service_username=lms_catalog_service_user

Then run the below command, this command will cause errors every time as it tries to cache programs from all sites that are added to your LMS sites model:

tutor local run lms ./manage.py lms cache_programs

If you don’t want the errors, then make use of an extra argument to the command .i.e. --domain. This argument will be equal to local.edly.io if you are running tutor local and local.edly.io:8000 if you are running tutor dev::

tutor local run lms ./manage.py lms cache_programs –domain=”local.edly.io” or tutor dev run lms ./manage.py lms cache_programs –domain=”local.edly.io:8000”

This last step should be performed every time you create new or make changes to existing programs.

Show Programs Tab

In order to show programs tab in the LMS dashboard, users will need to manually create an entry in Programs api config model in LMS Admin Panel. Go to http://local.edly.io/admin/programs/programsapiconfig/. Add Marketing path equal to /programs and enable it. Then Programs tab will be shown on the LMS where users can view their registered programs. It will show like as in below picture.

In the above image, User can see explore programs button which is pointing to http://localhost:8080/programs by default. This link does not exist. So, users can change this link to their custom build marketing site url to show all programs there. This can be done by modifying Site Configurations model in LMS Admin Panel. Go to http://local.edly.io/admin/site_configuration/siteconfiguration/. Open respective LMS site configuration and add below dict in site values field like below image:

"MKTG_URLS": {
    "ROOT": "https://custom-marketing-site-here.com"
}

Install extra requirements

In order to install extra requirements, use DISCOVERY_EXTRA_PIP_REQUIREMENTS and re-build the docker image.:

tutor config save \
  --set 'DISCOVERY_EXTRA_PIP_REQUIREMENTS=["git+https://github.com/myusername/myplugin"]'

Then, build the image, pointing to your fork if necessary:

tutor images build discovery

Debugging

To debug the course discovery service, you are encouraged to mount the course-discovery repo from the host in the development container:

tutor dev start –mount /path/to/course-discovery/ discovery

You can then access the development server at http://discovery.local.edly.io:8381. Feel free to add breakpoints (import pdb; pdb.set_trace()) anywhere in your source code to debug your application.

Once a local repository is mounted in the image, you will have to install nodejs dependencies and collect static assets:

tutor dev run discovery npm install --development
tutor dev run discovery make static.dev

Troubleshooting

This Tutor plugin is maintained by Muhammad Faraz Maqsood from Edly. Community support is available from the official Open edX forum. Do you need help with this plugin? See the troubleshooting section from the Tutor documentation.

Max retries exceeded with url

When running in production with HTTPS enabled, you may face this error during the init phase of tutor local launch:

` requests.exceptions.ConnectionError: HTTPSConnectionPool(host='<LMS_HOST>', port=443): Max retries exceeded with url: /api/courses/v1/courses/?page=1&page_size=10&username=discovery `

This error may be due to an incorrect DNS resolution of the LMS DNS record. With some cloud providers (for instance: [DigitalOcean](https://digitalocean.com/)) the /etc/hosts file on the host automatically contains the following entry:

127.0.1.1 <LMS HOST>

This entry may be present if you named your server with the LMS hostname.

License

This work is licensed under the terms of the GNU Affero General Public License (AGPL).