redis-caching-api

Redis caching API response

Simple caching example using Redis and some trivial API endpoint. Slow API responses are primary target when it comes to caching data. The same logic can be applied for any other, relational or not, database.

Prerequisites

• Node.js

$ apt install nodejs

• npm (Node Package Manager)

$ apt install npm

• Redis

$ apt install redis-server

Set Caching Evcition Policy and limit the available memory

Configure the Redis server to use to appropriate Eviction Policy, as well as to have the RAM limit.

nano /etc/redis/redis.conf

Change/Update the following configuration:

maxmemory 100mb
maxmemory-policy volatile-lru

Note that there is no difference between using the volatile-lru and allkeys-lru policy for this example since we're generating only expiration keys (keys with TTL set).

Trying out the application

Clone the repository to your computer:

$ git clone https://github.com/squirrelosopher/redis-caching/redis-caching-api

Navigate to the project folder:

$ cd redis-caching-api

Install dependencies:

$ npm install 

Running the example(s)

weather_api.js

The code is demonstrating HTTP response caching using Redis. It fetches the weather data for the given city either as the HTTP response (cache miss) or from the Redis server (cache hit). Timing is collected during the request execution and added to the returned JSON Object.

Ensure that the Redis is running, then run the:

$ node weather_api.js -n <city_name>

Since there is no cache entry, the program will retrieve the data via HTTP request.

$ node weather_api.js -n 'Belgrade'
{
  data: {
    coord: { lon: 20.4651, lat: 44.804 },
    weather: [ [Object] ],
    base: 'stations',
    main: {
      temp: 56.01,
      feels_like: 54.7,
      temp_min: 54.46,
      temp_max: 56.86,
      pressure: 1017,
      humidity: 72
    },
    visibility: 10000,
    wind: { speed: 17.27, deg: 130 },
    rain: { '1h': 0.21 },
    clouds: { all: 75 },
    dt: 1633552749,
    sys: {
      type: 2,
      id: 2039045,
      country: 'RS',
      sunrise: 1633495322,
      sunset: 1633536613
    },
    timezone: 7200,
    id: 792680,
    name: 'Belgrade',
    cod: 200
  },
  source: 'API',
  responseTime: '133ms'
}

The code will have a placed a copy of the entry in the Redis cache, so the next program execution will return data from the cache, resulting in much faster data retrieval.

$ node popular_movie.js -f '2000-01-01' -t '2020-01-01'
{
  data: {
    coord: { lon: 20.4651, lat: 44.804 },
    weather: [ [Object] ],
    base: 'stations',
    main: {
      temp: 56.01,
      feels_like: 54.7,
      temp_min: 54.46,
      temp_max: 56.86,
      pressure: 1017,
      humidity: 72
    },
    visibility: 10000,
    wind: { speed: 17.27, deg: 130 },
    rain: { '1h': 0.21 },
    clouds: { all: 75 },
    dt: 1633552749,
    sys: {
      type: 2,
      id: 2039045,
      country: 'RS',
      sunrise: 1633495322,
      sunset: 1633536613
    },
    timezone: 7200,
    id: 792680,
    name: 'Belgrade',
    cod: 200
  },
  source: 'cache',
  responseTime: '7ms'
}

help

You can always use the help parameter to get around with the program usage.

$ node weather_api.js --help

Check for Redis server

To check whether the server is running or not, you can use the command:

$ service redis-server status

The following commands are for starting, stopping and restarting the Redis server respectively:

$ service redis-server start

$ service redis-server stop

$ service redis-server restart

Change the Environment variables

If you have Redis set up and running on another host and/or port, or you want to specify a different API key, you can update it inside the .env file:

# If left out, environment variables have the default value set in constants.js file.

# Specifies port on which the Redis server is running
REDIS_PORT=6379

# Specifies host address on which the Redis server is running
REDIS_HOST=127.0.0.1

# Specifies the API key for the https://openweathermap.org/
WEATHER_API_KEY=aea864465325d76434ef9b28a2831f4e

API Information

The API used in the example for data fetching is OpenWeatherMap API. With free registration, the free pricing plan is offering a decent amount of available API calls per minute, enough for testing purposes.