Random Price Index Service

This is a test for people applying to the company with a mind to get involved with our technical infrastructure (Devops, Secops, SRE, Backend Dev, Data engineer etc).

Take as long as you need or you think is reasonable. You don't need to complete all the requirements if you are pushed for time, however your solution should give us enough to work with to confirm that you know what you're doing.

We have set up a google cloud project for your use in this test, and provided you the details and access via email (if you haven't received this let us know). The project is largely a blank slate, so feel free use whatever tooling or services you feel are appropriate to complete the task below, but try and avoid running up a crazy bill or exposing us to any potential security or legal problems.

Any configuration or documentation you might generate that you would like us to consider, please compress and return to us along with a copy of this git repository. If we could ask you not to make your solution public that would be appreciated.

Introduction

The Wayhome Random Price Index (or WRPI) is an exciting new (and importantly FICTIONAL) metric, similar in nature to the Retail Price Index (RPI) or the Consumer Price Index (CPI). Our data science team believe that WRPI is critical to our ability to buy houses, and as such should be integrated into our internal tooling post haste!

WRPI is a decimal number between 0 and 100. Unlike RPI and CPI that are only published once a month, WRPI can be updated much more frequently (around once per minute). Anyone consuming WRPI will need to know both the current value of WRPI and the time that it was generated at.

The Service

The backend team have worked with the data team to put together two python applications to satisfy this business function. See a service diagram below:

The changes to the web app are underway already, we just need to get these two applications running somewhere.

The Updater (updater.py)

The updater calculates the new value whenever it's run and stores the result in a redis database. The expectation is that this application will run approximately once per minute. The odd delay or duplication will not cause critical problems.

It takes the following configuration as environmental variables:

• REDIS_HOST: host for the redis database, defaults to localhost
• REDIS_DB: the number of the database to connect to, defaults to 1
• REDIS_PASS: password for the redis database, default to a blank string (or in other words: no password)
• REDIS_PORT: port for the redis database, defaults to 6379

The application sets the following keys in the database:

• wrpi-latest: the latest WRPI value
• wrpi-timestamp: the timestamp of the last calculation in ISO8601 format

The API (api.py)

This is a simple HTTP API that returns the current value of WRPI along with the timestamp it was generated at. It should be publicly exposed to the internet so that it might be accessed by our internal tooling (a browser based SPA). The application handles it's own authentication but you will need to configure and securely store the access password when you deploy the application.

The API should function something like this

GET 127.0.0.1/v0/wrpi HTTP/1.1
Host: 127.0.0.1
Authorization: Basic Ym9iOnBhc3N3b3Jk

200 OK
Content-Type: application/json
Content-Length: 123

{
    "latest_value": 69.420,
    "latest_calculation_at": "2021-02-20T18:06:47+0000"
}

Connecting via IP address is fine, the API doesn't need to be secured by SSL.

The application takes the following configuration as environmental variables:

• REDIS_HOST: host for the redis database, defaults to localhost
• REDIS_DB: the number of the database to connect to, defaults to 0
• REDIS_PASS: password for the redis database, default to a blank string (or in other words: no password)
• REDIS_PORT: port for the redis database, defaults to 6379
• APP_PASSWORD: the password users will need to enter in order to access the app. The application doesn't have a default but our users expect to use the same password we use for all our internal tools: security-goes-brrrrrr

Docker

The backend team have already packaged the two applications into docker images for you. You can find them attached to this github repository, or you can pull the images directly using the URI's below.

ghcr.io/wayhomeuk/random-price-index-service/updater:latest

ghcr.io/wayhomeuk/random-price-index-service/api:latest

Things that we would like to see

• We are currently a kubernetes shop and we would expect to see at least some demonstrated knowledge of kubernetes in your solution.
• A reproducible solution, i.e. configuration as code and clear documentation
• No sensitive data (such as kubernetes master passwords) being included in your code.
• We work with git for version control, please include your .git folder when you compress this folder and send it back to us. You should feel free to commit at any point in the process.

Stretch Goals

Don't feel like you need to go after any of these, providing a solution to the above will be enough for us to make an assessment. If you have time to kill and/or want to show off a bit, these are the things that we are interested in seeing.

• Monitoring and alerts.
• DNS and SSL for the API endpoint.
• Resource requests and limits.
• RBACs, IAM, permissions etc
• Networking magic (VPN, VPC, firewalls, etc)