- this allows us to send a notification to a provider by means of an API call
- This is in addition to the celery code.
- idea is that we can use this method to help speed up throughput by generating API traffic by node/lambda etc to supplement the celery code in times of high load.
This commit adds the `include_jobs` filter to the
`GET /services/…/notifications` endpoint. It defaults to `True` (ie show
all notifications) but makes it possible to only return notifications
created by _any_ API key.
This is so that we can show a log of all notifications sent through the
API in the admin app.
It does not expose this list to the public `GET /notifications` endpoint
because this would violate our rules about keys only being able to get
notifications created with keys of the same type.
- Thows a NoResultFound sqlalchemy exception
- Which causes a retry. This means we give it a few goes (5, max 5 hours) for the notification to appear.
- Should never happen, only if we get some task overlaps that are unusual that leads to tasks executed in an overlapping nature.
- Driven by the fact we won't know the type in the API call
- hence we need to load notification earlier , so pass it not the id through to the send task to avoid loading it twice.
- Aim to move the code that contacts providers into it's own module.
- Celery tasks now call this module to send to provider
- No exceptions caught in the new module. Celery tasks now use any exception to trigger a retry.
- tests moved about - new test directory for the new class, all tests from celery test module moved, excepting the retry logic.
- get all notifications by service
- template usage
- most recently used templates
Ensures that the dashboard shows no test key data. Supplements: https://github.com/alphagov/notifications-api/pull/677 which excludes CSV data. This branches from that so is dependant.
If you sign a token with a service ID that doesn’t exist (say, for
example, that you get service ID and API key mixed up) then you get
an error saying that “no API keys exist for the service”. This is wrong
because the service doesn’t even exist.
This commit adds:
- code to check if the service does exist
- a specific error message for this case
The check does mean an extra database call to look up the service.
However this only happens _after_ looping through all the API keys. So
it shouldn’t have a performance implication for anyone using a valid API
key.
We found that if the notifications were in created or pending they are not purged from notifications.
- New bulk update method to set all notificaitons with:
- a status = created|sending|pending to temporary-failure
- and is older then today minus SENDING_NOTIFICATIONS_TIMEOUT_PERIOD (in seconds)
- the scheduled task to timeout notifications use the new bulk update query.
- the task will be more efficient
We wrote to the queue as a performance optimisation , however dev found it confusing to not have immediate access to the notification as it may be perished some minutes later under periods of load. Additionally we had a couple of DB issues which led to us dropping notifications.
Pushing the DB write to earlier in the flow makes the system a little more robust in the early days, we may want to change this when the traffic increases.
