Skip to content

Latest commit

 

History

History
144 lines (92 loc) · 4.23 KB

README.md

File metadata and controls

144 lines (92 loc) · 4.23 KB
Raw

IP to Location

What is this?

This is an application that takes an IP and retrieves the location based on that IP. In other words, if you send your IP, the application respond you with your approximate location. To use the application, see the section "How to use it?" below.

The structure of the project was made following the below drawing:

Application diagram with the requested services

Apresentation

Running tests

Unitary

To run the tests, you'll just need npm and to install the dependencies:

npm i && npm run test:unitary

Integration

npm i && npm run test:integration

or npm test if you want to run both unitary and integration tests.

How to use it?

You have two ways to use this application, keep in mind that they have the same output:

Production

The recommended way, and only requires you to have node installed.

You just need to execute the silly-wizard.js:

node silly-wizard.js

And type a valid IPv4 or IPv6:

Prompt with input Prompt with input Prompt with output Prompt with broadcasted response

Since you are watching the output topic directly, the responses are broadcast-like, in other words, you'll see everyone else messages, and everyone else will see your messages.

Development

Recommended for development purpose. Must have docker and docker-compose installed.

Run the following command to get the app up a running. The .dev.env file has valid environment variables: 
docker compose --env-file .dev.env --profile full-app up

Then you will connect to the input and output topics. Open two terminal instances and in the first terminal, do the following to get access to the kafka service command line:

docker container exec -ti kafka bash

In that command line, create a consumer console connected to the development kafka broker and the 'location_output' topic:

kafka-console-consumer --bootstrap-server kafka:9092 --topic location_output

In the other terminal instance, connect to the kafka again:

docker container exec -ti kafka bash

But now, create a producer connected to the 'location_input':

kafka-console-producer --bootstrap-server kafka:9092 --topic location_input

As soon as the console is ready, you can produce messages with the following schema:

{"clientId": "your_client_id", "timestamp": 123, "ip": "insert_ipv4_or_ipv6"}

And then check the consumer console. The location must be there :)

Direct connection

If you want to type your keys a bit, you can connect to the production broker at ec2-15-228-13-15.sa-east-1.compute.amazonaws.com:9094. The topics are 'location_input' and 'location_output'.

Decisions through development phase

  • Currently, the topics only accept and producer messages that are a valid JSON with 'clientId', 'timestamp' and 'ip'. Unfortunately, the client doesn't have ways to know what is the schema of the message, unless they read this README. The way I could solve this problem would be with a schema registry. In the draw I put in the begining of the file would have an extra service between the input and output topic.

  • One problem that could be solved with extra service was the ignored messages. Currently, the application invalid messages that is if the message is not a valid JSON, if ipstack doesn't return a valid answer, etc... If a fail case occurs, the application just forget about the message. Which I think it's not ideal, I would solve this issue with an extra topic called 'invalid_messages' that would be a retry queue or maybe a dead letter queue for minimum analytics.