Designed and developed a Java application that acts as an HTTP proxy for incoming requests. The application exposes an API endpoint that can receive an ID and another URL endpoint and then calls the endpoint provided as argument with the number of unique IDs received so far. The number of unique IDs received is also periodically written to multiple sinks (e.g., log file, Redis) using the Spring Scheduler mechanism. Tested the load of the application using JMeter.

Tech: Java, Spring, Spring Boot, Lombok, Redis, JMeter, Docker

GET /api/accept?id=123&endpoint=https://www.google.com

• id = represents the id of the request
• endpoint = represents an HTTP endpoint that should be called by the application

• Spring Boot
• Lombok
• Redis
  • for de-duplicating ids: uses the set data structure for storing unique requests ids
  • as a streaming platform: queue for sending the number of unique requests in the last minute
• JMeter
  • for load testing

If you want to test the application you can use the following Docker Redis image: https://hub.docker.com/_/redis

The application is structured as follows:

• configuration: configuration classes
• controller: controller classes for exposing the HTTP API
• service: business logic for controllers
• model: data classes
• unique.requests.producer: business logic for periodically writing unique ids to redis and log files
• deduplication: business logic for storing unique ids using Redis and Sets

• In order to run the application you need a Redis instance to connect to.

• Update the application.properties for the app with the Redis host and port.

• ./gradlew bootRun

./gradlew bootJar

Update the properties in application.properties

• api base path
• server port
• Redis host and port
• Redis streaming queue
• Redis set name