Pre-load the course overviews attached to CourseEnrollments on the
Student Dashboard, if possible. This will only grab the CourseOverviews
that already exist, and will not generate new ones. Any missing
CourseOverviews fall back to the lazily-created one-at-a-time behavior
they've always had. That's mostly because I wanted to optimize for the
common case in the least invasive way possible, and I don't want to get
caught up in locking issues.
There are a number of Django Signals that are on the modulestore's
SignalHandler class, such as SignalHandler.course_published. These
signals can trigger very expensive processes to occur, such as course
overview or block structures generation. Most of the time, the test
author doesn't care about these side-effects.
This commit does a few things:
* Converts the signals on SignalHandler to be instances of a new
SwitchedSignal class, that allows signal sending to be disabled.
* Creates a SignalIsolationMixin helper similar in spirit to the
CacheIsolationMixin, and adds it to the ModuleStoreIsolationMixin
(and thus to ModuleStoreTestCase and SharedModuleStoreTestCase).
* Converts our various tests to use this new mechanism. In some cases,
this means adjusting query counts downwards because they no longer
have to account for publishing listener actions.
Modulestore generated signals are now muted by default during test runs.
Calls to send() them will result in no-ops. You can choose to enable
specific signals for a given subclass of ModuleStoreTestCase or
SharedModuleStoreTestCase by specifying an ENABLED_SIGNALS class
attribute, like the following example:
from xmodule.modulestore.tests.django_utils import ModuleStoreTestCase
class MyPublishTestCase(ModuleStoreTestCase):
ENABLED_SIGNALS = ['course_published', 'pre_publish']
You should take great care when disabling signals outside of a
ModuleStoreTestCase or SharedModuleStoreTestCase, since they can leak
out into other tests. Be sure to always clean up, and never disable
signals outside of testing. Because signals are essentially process
globals, it can have a lot of unpleasant side-effects if we start
mucking around with them during live requests.
Overall, this change has cut the total test execution time for
edx-platform by a bit over a third, though we still spend a lot in
pre-test setup during our test builds.
[PERF-413]
We need to be able to track which site a given user account was created on. This
change will create a UserAttribute model with a key of 'created_on_site' and a value
containing the domain of the site on which the user accounted was created.
WL-977
This adds functions to the catalog utils which munge data
collected from the program endpoint and LMS database to construct
data structures that are ready for use by view and templates
related to the Open EdX marketing pages that live in LMS.
WL-766
Updates Mako and Underscore templates as well as Backbone models and views so they work with catalog programs. Removes all remaining response munging from the back end.
ECOM-4422
This commit contains back end changes necessary to load programs from the catalog in all contexts. The existing program munging utility is applied as late as possible to avoid conflating this work with changes to the front end; those will be made separately.
ECOM-4422
We should not be using custom signing keys for each service at this time. We may want to return to this strategy in the future; but, this is not the direction any of our other services are going in.
ECOM-6541
This command is used to backfill missing program certificates. Formerly, it used the programs service. It now uses the catalog service exclusively.
ECOM-4422
All programs are now managed using the Django admin on the catalog service. This is the first in a series of commits removing use of the old programs service from edx-platform.
ECOM-4422
This command was updated to request programs from the catalog, but it wasn't changed to deal with the new response structure. I'm updating all code in the LMS to deal with catalog service responses separately.
ECOM-6900.
When crawlers like edX-downloader make requests on courseware, they are
often concurrently loading many units in the same sequence. This causes
contention for the rows in courseware_studentmodule that store the
student's state for various XBlocks/XModules, most notably for the
sequence, chapter, and course -- all of which record and update user
position information when loaded.
It would be nice if we could actually remove these writes altogether
and come up with a cleaner way of keeping track of the user's position.
In general, GETs should be side-effect free. However, any such change
would break backwards compatibility, and would require close
coordination with research teams to make sure they weren't negatively
affected.
This commit identifies crawlers by user agent (CrawlersConfig model),
and blocks student state writes if a crawler is detected. FieldDataCache
writes simply become no-ops. It doesn't actually alter the rendering
of the courseware in any way -- the main impact is that the blocks
won't record your most recent position, which is meaningless for
crawlers anyway.
This can also be used as a building block for other policy we want to
define around crawlers. We just have to be mindful that this only works
with "nice" crawlers who are honest in their user agents, and that
significantly more sophisticated (and costly) measures would be
necessary to prevent crawlers that try to be even trivially sneaky.
[PERF-403]
The programs service has no concept of run publication status. The catalog does, and many runs returned by the programs service are marked as unpublished in the catalog. To maintain existing behavior and continue displaying these runs as part of their respective programs, we need to ignore their publication status.
ECOM-6925
The system checks require database access, which is not available when building Docker images. This relaxes the check, allowing the command to execute without a database.
ECOM-6634
