Describe the bug

Upon Pull Request opening, the CONTRIBUTING link within it is clearly broken for the absence of a correct path.

To Reproduce

1. Open a Pull Request
2. Click on the template-provided CONTRIBUTING link
3. See "Not Found"

Expected behaviour

The link should bring you to https://github.com/immuni-app/immuni-backend-exposure-ingestion/blob/master/CONTRIBUTING.md

Issue content

After running the docker compose command as described in the documentation I get the following error: Need service name for --build-arg option, command below:

test@test-VirtualBox:~/immuni/immuni-backend-exposure-ingestion/docker$ docker-compose build     --build-arg GIT_BRANCH=$(git rev-parse --abbrev-ref HEAD)     --build-arg GIT_SHA=$(git rev-parse --verify HEAD)     --build-arg GIT_TAG=$(git tag --points-at HEAD | cat)     --build-arg BUILD_DATE=$(date -u +"%Y-%m-%dT%H:%M:%SZ")
ERROR: Need service name for --build-arg option

Docker version:

test@test-VirtualBox:~/immuni/immuni-backend-exposure-ingestion/docker$ sudo docker version
Client:
 Version:           19.03.6
 API version:       1.40
 Go version:        go1.12.17
 Git commit:        369ce74a3c
 Built:             Fri Feb 28 23:45:43 2020
 OS/Arch:           linux/amd64
 Experimental:      false

There is a typo in CONTRIBUTING.md

security and quality standards in the development tools described in should be security and quality standards using the development tools described in

The CONTRIBUTING documentation describes checks that do run upon Pull Request opening, yet they are not currently present.

This is the link to the section.

The CODEOWNERS file should be structured as follows:

# These owners will be the default owners for everything in
# the repo. Unless a later match takes precedence,
# @global-owner1 and @global-owner2 will be requested for
# review when someone opens a pull request.
*       @global-owner1 @global-owner2

The MAX_ALLOWED_CLIENT_SKEW_IN_SECONDS configuration variable found in immuni_exposure_ingestion/core/config.py is unused.

It's minor, but it should be removed.

Describe the bug

Endpoint v1/ingestion/upload returns status code 500 if the teks parameter is an empty list.

To Reproduce

1. cd docker
2. docker-compose build
3. docker-compose up

curl --location --request POST 'localhost:5000v1/ingestion/upload' \
--header 'Immuni-Dummy-Data: 0' \
--header 'Immuni-Client-Clock: 1589903340' \
--header 'Content-Type: application/json; charset=utf-8' \
--header 'Authorization: Bearer 4ec50b1f75ef4ad97521f5b3610cee605595266b7e2c42e6ce72eadff067c108' \
--data-raw '{
  "exposure_detection_summaries": [
    {
      "attenuation_durations": [
        300,
        0,
        0
      ],
      "date": "2020-05-27",
      "days_since_last_exposure": 1,
      "exposure_info": [
        {
          "attenuation_durations": [
            300,
            0,
            0
          ],
          "attenuation_value": 45,
          "date": "2020-05-16",
          "duration": 5,
          "total_risk_score": 4,
          "transmission_risk_level": 1
        }
      ],
      "matched_key_count": 2,
      "maximum_risk_score": 4
    }
  ],
  "padding": "S0meP4Dd1nG",
  "province": "AG",
  "teks": []
}'

Expected behaviour

The server should return a more explanatory error with status code 400.

First of all, thank you for all the effort you put in this project, your care for sensitive topics has convinced even the paranoid people (like myself) to use the app. Kudos.

I’m a bit concerned about the concurrency model, you have chosen a really fast async framework (Sanic) running with the AsyncIO event loop but you are using incompatible network libraries, most notably MongoEngine, which is a mongoDB ODM on top of pymongo. Pymongo does not support AsyncIO

Let’s take a look at this chunk of code

Moreover, here

• You are explicitly avoiding concurrency with a distributed lock

  immuni-backend-exposure-ingestion/immuni_exposure_ingestion/tasks/process_uploads.py

  Line 56 in cbd2aca

  async with lock_concurrency("process_uploads"):

• you are running blocking code

  immuni-backend-exposure-ingestion/immuni_exposure_ingestion/tasks/process_uploads.py

  Line 78 in cbd2aca

  uploads = Upload.to_process()

Why bother with all the concurrency "inception" (Celery -> AsyncIO evt loop -> coroutine) then, when what you want achieve is a serial execution? There is no need of concurrency here. What about a celery task that runs synchronous code?
One could also argue that even Celery is an overkill when the need is to launch a cron job, there are a bunch of tools that allows to keep control of task scheduling on the application layer, without all the celery machinery, APScheduler for instance.

Cheers