Devstack is the local Docker-based environment for developing in the Open edX platform. Use it to get up and running quickly with Open edX services.
Documentation is on Read the Docs. Code repository is on GitHub.
The Devstack runs as multiple containers with Docker Compose at its core.
A Devstack installation includes the following Open edX components by default:
- The Learning Management System (LMS)
- The Learning micro-frontend (A.K.A the new Courseware experience)
- Open Response Assessments (ORA2), among other LMS plug-ins.
- Open edX Studio
- Discussion Forums
- E-Commerce
- Credentials
- Notes
- Course Discovery
- Open edX Search
- A demonstration Open edX course
- The Publisher and Gradebook micro-frontends
It also includes the following extra components:
- XQueue
- The Program Console micro-frontend
- The Library Authoring micro-frontend
- edX Registrar service.
- The course-authoring micro-frontend
- The enhanced staff grader (ora-grading) micro-frontend
Table of Contents:
There are a number of places to get help, including mailing lists and real-time chat. Please choose an appropriate venue for your question. This helps ensure that you get good prompt advice, and keeps discussion focused. For details of your options, see the Community pages.
- See the most common development workflow (after you've finished Getting Started).
- See the Devstack Interface
- See some helpful troubleshooting tips.
- See the Frequently Asked Questions.
- Or learn about testing and debugging your code in devstack.
- If you get confused about any of the terms used in these docs, see edX Glossary
You can also browse all the documentation in Read the Docs.
NOTE: LMS is now using MySql 5.7 by default. You have to run make dev.pull.lms and make dev.provision.lms (more details in Getting Started) to fetch latest images and reprovision local copies of databases in order for an existing devstack setup to keep working.
You will need to have the following installed:
- make
- Python 3.8
- Docker
This project requires Docker 17.06+ CE. We recommend Docker Stable, but Docker Edge should work as well.
NOTE: Switching between Docker Stable and Docker Edge will remove all images and settings. Don't forget to restore your memory setting and be prepared to provision.
For macOS users, please use Docker for Mac. Previous Mac-based tools (e.g. boot2docker) are not supported. Please be aware that the licensing terms for Docker for Mac (aka Docker Desktop) may mean that it is no longer free for your organization's use.
Since a Docker-based devstack runs many containers, you should configure Docker with a sufficient amount of resources. We find that configuring Docker for Mac with a minimum of 2 CPUs, 8GB of memory, and a disk image size of 96GB does work.
Docker for Windows may work but has not been tested and is not supported.
If you are using Linux, use the overlay2 storage driver, kernel version
4.0+ and not overlay. To check which storage driver your
docker-daemon uses, run the following command.
docker info | grep -i 'storage driver'You should run all make commands described below on your local machinge, not
from within a Virtual Machine, as these commands are meant to stand up a VM-like environment using
Docker containers.
However, you may want to run the make commands from within a Python 3 virtual
environment, as described in Getting Started. This will keep the Python packages required for Devstack separate from
the ones installed globally on your system.
The default devstack services can be run by following the steps below.
Note: This will set up a large number of services, more than you are likely to need to work with, but that's only necessary for first-time provisioning. See Service List and the most common development workflow for how to run and update devstack with just the services you need, rather than the large-and-slow default set.
Install the requirements inside of a Python virtualenv.
make requirements
This will install docker-compose and other utilities into your virtualenv.
The Docker Compose file mounts a host volume for each service's executing code. The host directory defaults to be a sibling of this directory. For example, if this repo is cloned to
~/workspace/devstack, host volumes will be expected in~/workspace/course-discovery,~/workspace/ecommerce, etc. These repos can be cloned with the command below.make dev.clone # or, `make dev.clone.https` if you don't have SSH keys set up.You may customize where the local repositories are found by setting the
DEVSTACK_WORKSPACEenvironment variable.(macOS only) Share the cloned service directories in Docker, using Docker -> Preferences -> File Sharing in the Docker menu.
Pull any changes made to the various images on which the devstack depends.
make dev.pull.large-and-slow
Note - If you are setting up devstack to develop on Open edx named releases, see this document on developing on named releases before following this step 3.
Run the provision command, if you haven't already, to configure the various services with superusers (for development without the auth service) and tenants (for multi-tenancy).
NOTE: When running the provision command, databases for ecommerce and edxapp will be dropped and recreated.
The username and password for the superusers are both
edx. You can access the services directly via Django admin at the/admin/path, or login via single sign-on at/#/.Default:
make dev.provision
This is expected to take a while, produce a lot of output from a bunch of steps, and finally end with
Provisioning complete!Start the desired services. This command will mount the repositories under the
DEVSTACK_WORKSPACEdirectory.NOTE: it may take up to 60 seconds for the LMS to start, even after the
dev.up.*command outputsdone.Default:
make dev.up.large-and-slow
To stop a service, use make dev.stop.<service>, and to both stop it
and remove the container (along with any changes you have made
to the filesystem in the container) use make dev.down.<service>.
After the services have started, if you need shell access to one of the
services, run make dev.shell.<service>. For example to access the
Catalog/Course Discovery Service, you can run:
make dev.shell.discoveryTo see logs from containers running in detached mode, you can either use "Kitematic" (available from the "Docker for Mac" menu), or by running the following:
make dev.logsTo view the logs of a specific service container run make dev.logs.<service>.
For example, to access the logs for Ecommerce, you can run:
make dev.logs.ecommerceFor information on the supported make commands, you can run:
make helpDevstack collects some basic usage metrics to help gain a better understanding of how devstack is used and to surface any potential issues on local devstack environments. To learn more, read 0003-usage-metrics.rst ADR.
This data collection is behind a consent flag, so please help devstack's maintainers by enabling metrics collection by running the following:
make metrics-opt-inNow that you're up and running, read about the most common development workflow.
The provisioning script creates a Django superuser for every service.
Email: edx@example.com Username: edx Password: edx
The LMS also includes demo accounts. The passwords for each of these accounts
is edx.
Account Description staff@example.comAn LMS and Studio user with course creation and editing permissions. This user is a course team member with the Admin role, which gives rights to work with the demonstration course in Studio, the LMS, and Insights. verified@example.comA student account that you can use to access the LMS for testing verified certificates. audit@example.comA student account that you can use to access the LMS for testing course auditing. honor@example.comA student account that you can use to access the LMS for testing honor code certificates.
These are the edX services that Devstack can provision, pull, run, attach to, etc.
Each service is accessible at localhost on a specific port.
The table below provides links to the homepage, API root, or API docs of each service,
as well as links to the repository where each service's code lives.
Most developers will be best served by working with specific combinations of these services, for example make dev.pull.studio or make dev.up.ecommerce. These will pull in dependencies as needed—starting ecommerce will also start lms, and lms will pull in forums, discovery, and others. If you need multiple, they can be listed like make dev.up.studio+ecommerce. After the service table below there is a list of some common combinations.
Instead of a service name or list, you can also run commands like make dev.provision / make dev.pull.large-and-slow / make dev.up.large-and-slow. This is a larger list than most people will need for most of their work, and includes all of the services marked "Default" in the below table. (Some of these targets use large-and-slow in their name as a warning; others may be changed to use this over time.) However, you can change this list by modifying the DEFAULT_SERVICES option as described in the Advanced Configuration Options section.
Some common service combinations include:
lms: LMS, along with dependenciesforum,discovery,Authnand some databasesecommerce: Ecommerce, but also LMS as a dependency (for auth)studio+credentials: Services can be combined to affect both at once
Currently, some containers rely on Elasticsearch 7 and some rely on Elasticsearch 1.5. This is because services are in the process of being upgraded to Elasticsearch 7, but not all of them support Elasticsearch 7 yet. As we complete these migrations, we will update the dependencies of these containers.
The file options.mk sets several configuration options to default values.
For example DEVSTACK_WORKSPACE (the folder where your Git repos are expected to be)
is set to this directory's parent directory by default,
and DEFAULT_SERVICES (the list of services that are provisioned and run by default)
is set to a fairly long list of services out of the box.
For more detail, refer to the comments in the file itself.
If you're feeling brave, you can create an git-ignored overrides file called
options.local.mk in the same directory and set your own values. In general,
it's good to bring down containers before changing any settings.
The COMPOSE_PROJECT_NAME variable is used to define Docker namespaced volumes
and network based on this value, so changing it will give you a separate set of databases.
This is handled for you automatically by setting the OPENEDX_RELEASE environment variable in options.mk
(e.g. COMPOSE_PROJECT_NAME=devstack-juniper.master. Should you want to manually override this, edit the options.local.mk in the root of this repo and create the file if it does not exist. Change the devstack project name by adding the following line:
# Example: COMPOSE_PROJECT_NAME=secondarydevstack COMPOSE_PROJECT_NAME=<your-alternate-devstack-name>
As a specific example, if OPENEDX_RELEASE is set in your environment as juniper.master, then COMPOSE_PROJECT_NAME will default to devstack-juniper.master instead of devstack.