ML - MAGNETO

The current IP is 52.201.56.84. But I will stop the instance during this weekend. But in any case, you can run this exercise locally by just typing:

make

You need Docker Compose installed in your Linux system.

Level 1

The main function that calculates if the DNA sequence corresponds to a human or a mutant can be found in src/ml/mutant/mutant.go with the IsMutant(input []string) (bool, error) signature.

The original input []string is transformed into an NXM array of interconnected Nodes. Each node is connected with the Node.nodes variable. And to prevent re-testing combinations the Node.dicarted variable is used. The code has the concept of directions, and is a way of describing if the algorithm took the path to the right, left, up or down relative to the x element:

0  1  2
7  X  3
6  5  4

I'm sure there is some cool graph theory algorithm that could greatly simplify what I have done. If you know how to improve it, tell me.

Level 2

src/ml/main.go has the standard go webserver. If you run make you should be able to interact with the API:

• /mutant
• /stats (it updates every 5seconds)

Each endpoint sets Cache-Control. It may be useful if the server is behind a reverse proxy, or maybe a CDN. In this way.

Level 3

src/stats/stats.go stores the results of the DNA checks in an SQLite database. It uses the getCachedStats function to cache results. I created that "eventually consistent" cache function because I thought that was what I was expected to do, so that constant COUNTs queries are avoided.

100 to 1M QPS

It's a little bit hard to assume this program could support 1M qps, but because the Cache-Control headers are set, a CDN like AWS CloudFront could be configured to pass POST requests while honoring the Cache-Control.