The data flow of other bits of our application looks like this:
```
API (returns JSON)
⬇
API client (returns a built in type, usually `dict`)
⬇
Model (returns an instance, eg of type `Service`)
⬇
View (returns HTML)
```
The user API client was architected weirdly, in that it returned a model
directly, like this:
```
API (returns JSON)
⬇
API client (returns a model, of type `User`, `InvitedUser`, etc)
⬇
View (returns HTML)
```
This mixing of different layers of the application is bad because it
makes it hard to write model code that doesn’t have circular
dependencies. As our application gets more complicated we will be
relying more on models to manage this complexity, so we should make it
easy, not hard to write them.
It also means that most of our mocking was of the User model, not just
the underlying JSON. So it would have been easy to introduce subtle bugs
to the user model, because it wasn’t being comprehensively tested. A lot
of the changed lines of code in this commit mean changing the tests to
mock only the JSON, which means that the model layer gets implicitly
tested.
For those reasons this commit changes the user API client to return
JSON, not an instance of `User` or other models.
when clients are defined in app/__init__.py, it increases the chance of
cyclical imports. By moving module level client singletons out to a
separate extensions file, we stop cyclical imports, but keep the same
code flow - the clients are still initialised in `create_app` in
`__init__.py`.
The redis client in particular is no longer separate - previously redis
was set up on the `NotifyAdminAPIClient` base class, but now there's one
singleton in `app.extensions`. This was done so that we can access redis
from outside of the existing clients.
This removes some code which is duplicative and obscure (ie it’s not
very clear why we do `"a" * 73` even though there is a Very Good Reason
for doing so).
This is better than just keying into the JSON because it means you get
an exception straight away when looking up a key that doesn’t exist
(which via mocking you could ordinarily miss).
Most of the time spent by the admin app to generate a page is spent
waiting for the API. This is slow for three reasons:
1. Talking to the API means going out to the internet, then through
nginx, the Flask app, SQLAlchemy, down to the database, and then
serialising the result to JSON and making it into a HTTP response
2. Each call to the API is synchronous, therefore if a page needs 3 API
calls to render then the second API call won’t be made until the
first has finished, and the third won’t start until the second has
finished
3. Every request for a service page in the admin app makes a minimum
of two requests to the API (`GET /service/…` and `GET /user/…`)
Hitting the database will always be the slowest part of an app like
Notify. But this slowness is exacerbated by 2. and 3. Conversely every
speedup made to 1. is multiplied by 2. and 3.
So this pull request aims to make 1. a _lot_ faster by taking nginx,
Flask, SQLAlchemy and the database out of the equation. It replaces them
with Redis, which as an in-memory key/value store is a lot faster than
Postgres. There is still the overhead of going across the network to
talk to Redis, but the net improvement is vast.
This commit only caches the `GET /service` response, but is written in
such a way that we can easily expand to caching other responses down the
line.
The tradeoff here is that our code is more complex, and we risk
introducing edge cases where a cache becomes stale. The mitigations
against this are:
- invalidating all caches after 24h so a stale cache doesn’t remain
around indefinitely
- being careful when we add new stuff to the service response
---
Some indicative numbers, based on:
- `GET http://localhost:6012/services/<service_id>/template/<template_id>`
- with the admin app running locally
- talking to Redis running locally
- also talking to the API running locally, itself talking to a local
Postgres instance
- times measured with Chrome web inspector, average of 10 requests
╲ | No cache | Cache service | Cache service and user | Cache service, user and template
-- | -- | -- | -- | --
**Request time** | 136ms | 97ms | 73ms | 37ms
**Improvement** | 0% | 41% | 88% | 265%
---
Estimates of how much storage this requires:
- Services: 1,942 on production × 2kb = 4Mb
- Users: 4,534 on production × 2kb = 9Mb
- Templates: 7,079 on production × 4kb = 28Mb
Currently requests to the API made from the admin app are going from
PaaS admin app to the nginx router ELB, which then routes them back
to the api app on PaaS.
This makes sense for external requests, but for requests made from
the admin app we could skip nginx and go directly to the api PaaS
host, which should reduce load on the nginx instances and
potentially reduce latency of the api requests.
API apps on PaaS are checking the X-Custom-Forwarder header (which
is set by nginx on proxy_pass requests) to only allow requests going
through the proxy.
This adds the custom header to the API client requests, so that they
can pass that header check without going through nginx.
in the NotifyAdminAPIClient, which all api traffic goes through, return
403 for any stateful requests (post, put and delete), if the following
criteria have been met:
* a current_service is set
(this prevents checks being carried out on non-service related
updates, eg editing user details)
* the service is not active
* the current user is not a platform admin
so platform admins can still update anything.
Note: Without any specific error handling, the user will see a generic
403 page. This is fine, probably - it's a relatively niche case that
you'll be editing a service you can't get to anyway
in the NotifyAdminAPIClient, which all api traffic goes through, return
403 for any stateful requests (post, put and delete), if the following
criteria have been met:
* a current_service is set
(this prevents checks being carried out on non-service related
updates, eg editing user details)
* the service is not active
* the current user is not a platform admin
so platform admins can still update anything.
Note: Without any specific error handling, the user will see a generic
403 page. This is fine, probably - it's a relatively niche case that
you'll be editing a service you can't get to anyway
- This allows us to set a custom header for admin calls only (not needed in client calls)
- Adds request-id from Middleware to the API call to ensure the API logs against the same request ID.
This PR changes the flow to change an email address.
Once the user enter their password, they are told "Check your email".
An email has been sent to them containing a link to notify which contains an encrypted token.
The encrypted token contains the user id and new email address. Once the link is clicked the user's email address is updated to the new email address.
They are redirected to the /user-profile page.
Also in this commit is an update from flask.ext.login to flask_login.
Mutating dictionaries is gross and doesn’t work as you’d expect. Better
to have the function return a new dictionary instead.
Means we can be explicit that `created_by` is one of the allowed params
when updating a service.
The post register endpoint will send a random 5 digit code via sms and another via email.
If either code fails to send, the user will not be created and the person can register again.
The codes are saved to the session cookie, and expire in 1 hour.
Another iteration of this story will save the codes to a database.
