Atea

I spent 11 hours on an assignment to get an interview with Atea in Norway. I have worked with IT since forever. You have used things I have made.

Feel free to check out my CV. Read the original assignment below.

Introduction

Their response was:

Bad design skills.
The code is mostly not readable.
Bad code quality.

Here is my application, as provided to them before their response:

Task 1

This solves task 1 of the Atea assignment.

Some solution information:

Uses Azure Functions with a TimerTrigger to run every 1 minutes
Two separate projects, one for the Azure Functions and one for the REST API
Both implemented with .NET 6 LTS
REST API is minimal API
REST API implements paging with continuation token (not very well tested)
The code is commented a little more than I would normally do, but I wanted to make it easier to understand my reasoning
No automatic testing is implemented

Some issues I ran into:

I solved this on a relatively clean computer, takes time to setup tooling for local Azure development
I didn't have experience with Azurite, so I had to spend some time figuring out how to use it compared to the Azure Storage Emulator
- I ended up using Azurite, but running from CLI instead of implicitly as part of the Function. I am sure this can be done better, but I didn't have time to figure it out

Time spent: Approximately 4-5 hours. I spent a lot of time setting up the tooling and figuring out how to use Azurite. Most of the code is trivial, but I wanted to make sure it was well commented and easy to understand in this special case where I am trying to show my skills.

Task 2

Some solution information:

Uses ASP.NET with minimal API
Registers a BackgroundService that runs every 1 minutes using a PeriodicTimer
Stores cities in configuration using the options-pattern
Uses fan-out parallelization for fetching weather data using HTTP
Uses System.Text.Json for JSON serialization
Uses EF Core with SQLite for storing weather data
Uses Meteo as weather provider (might be a bad choice, but it was the first suitable one I found)
I took some liberties and combined the datasets into one table, since I didn't see any reason to have them separate. They can be toggled by clicking in the header of the chart

Some issues I ran into:

I have not used SQLite with EF Core before, so I had to spend some time figuring out how to use it
The requirements are very vague, so I had to make some assumptions

Did not finish:

In the alotted time, I couldn't make the chart respond to clicks in a meaningful way. I would have liked to have the chart show the data for the selected city, but I didn't have time to implement it

Time spent: Probably around 5-6 hours.

Other improvements

There is use of HttpClient, could conceivably have used something like Flurl instead
Issues with the Meteo API:
- It does not return cloud cover as part of current_weather. Seeing how it wasn't asked to chart it, I didn't include quering and parsing it in the model. This is a relatively easy fix, but I didn't have time to do it
- It seems like current_weather is not updated very often, so the data is not very accurate. I have not found a way to get more accurate data out of Meteo. This results in the chart being very flat and a lot of the data being the same in the database
I didn't know how to visualize the timestamp in the chart. I suppose it could be part of the tooltip, but I would love to have some feedback from product owner on it
Probably 100 other things :)

Total time spent: Perhaps 10-11 hours.

How to run

Just clone it, open it in Visual Studio 2022 and run it. It should just work. You need to run azurite, I run it like this:

azurite -s -l c:\azurite -d c:\azurite\debug.log

You also probably need to initialize the database for WeatherPlot:

dotnet ef database update

This will run the initial migration and create the database.

The original assignment

Original assignment

We have made it a general practice in our recruitment (FTE or Consultants) to conduct a screening process through a little homework that demonstrates the skill level they possess in the specific areas that are of particular importance to us.

The Process & instructions:

Timeframe: April 12th

Please, describe each task separately:

How much time it took for you to complete it (hours from start to finish);

Some challenges/ideas/comments on each task;

After receiving the completed tasks, we will review them and book you a meeting where we will share with our conclusions before we move forward

Technical assignments:

First task

Must use:

Azure Function (Cloud/Local)
Azure Storage (Cloud /Local storage emulator)

a. Table

b. Blob

• .Net Core 6

Achieve:

• Every minute, fetch data from https://api.publicapis.org/random?auth=null and store success/failure attempt log in the table and full payload in the blob.

• Create a GET API call to list all logs for the specific time period (from/to)

• Create a GET API call to fetch a payload from blob for the specific log entry

• Publish code on GitHub (public)

Second task

Must use:

• ASP.NET CORE MVC (6)

• C#

• JavaScript or Typescript

Achieve:

Using any public weather API receive data (country, city, temperature, clouds, wind speed) from at least 10 cities in 5 countries

with periodical update 1/min,

store this data in the database

and show the 2 graphs:

min temperature (Country\City\Temperature\Last update time)
highest wind speed (Country\City\Wind Speed\Last update time)
temperature & wind speed trend for last 2 hours on click for both previous graphs

COCPORN/atea