[Node.js] Level 7. Persisting Data
Simple Redis Commands
Let's start practicing using the redis key-value store from our node application.
Require the redis
module and assign it to a variable called redis
.
var redis = require('redis');
Create a redis
client and assign it to a variable called client
.
var client = redis.createClient()
On the client
, set the name
property to your name.
client.set('name', 'Answer');
var redis = require('redis'), client = redis.createClient(); client.set('name', 'Answer');
Get A Key
We have already stored a value in the question key. Use the redis client to issue a get
command to redis to retrieve and then log the value.
Use the redis client to issue a get
command using the 'question'
key to retrieve a value. Remember, the get
function takes a callback which expects two arguments, error
and data
.
Log the value retrieved with console.log
.
var redis = require('redis'); var client = redis.createClient(); client.get('question', function(err, data){ console.log(data); });
Working With Lists 1
As we saw in the video, redis can do more than just simple key-value pairs. We are going to be using redis' LISTS later to add persistence to our live-moderation app, so let's practice using them now
Using the redis client's lpush
command, insert question1
into thequestions
list. Then, console.log
the result you receive. Remember, thelpush
function takes a callback as its last argument, which expects anderror
and value
to be passed as arguments.
client.lpush('questions', question1, function(err, data){ console.log(data); });
Using the redis client's lpush
command, insert question2
into the questions list. Then console.log
the result you receive.
var redis = require('redis'); var client = redis.createClient(); var question1 = "Where is the dog?"; var question2 = "Where is the cat?"; client.lpush('questions', question1, function(err, data){ console.log(data); }); client.lpush('questions', question2, function(err, data){ console.log(data); });
Working With Lists 2
Now that we have seeded the questions
list, use the lrange()
command to return all of the items and log them.
Use the lrange()
command to return all of the items from the questions
key.
Now that we have called lrange()
, use console.log
to log the result from redis.
var redis = require('redis'); var client = redis.createClient(); client.lrange('questions', 0, -1, function(err, data){ console.log(data); });
Persisting Questions
Let's go back to our live-moderation app and add some persistence, first to the questions people ask.
Use the lpush
command to add new questions to the list namedquestions
. Do this inside the listener for the 'question'
event.
redisClient.lpush('questions', question);
var express = require('express'); var app = express(); var server = require('http').createServer(app); var socket = require('socket.io'); var io = socket.listen(server); var redis = require('redis'); var redisClient = redis.createClient(); io.sockets.on('connection', function(client) { client.on('answer', function(question, answer) { client.broadcast.emit('answer', question, answer); }); client.on('question', function(question) { if(!client.question_asked) { client.question_asked = true; client.broadcast.emit('question', question); // add the question to the list here redisClient.lpush('questions', question); } }); });
Emitting Stored Questions
Now that we have questions stored in redis, let's emit them whenever a new client connects to the server through socket.io.
Use the lrange
command to retrieve a list of questions that represent thequestions
list within redis.
Inside of the lrange
callback, use a forEach
loop to iterate through thequestions
and emit()
each question
to the client. Remember, don't usebroadcast.emit
because we only want to send the questions to the client that is connecting to the server.
redisClient.lrange('questions', 0, -1, function(err, questions){ questions.forEach(function(question){ client.emit("question", question); }); });
var express = require('express'); var app = express(); var server = require('http').createServer(app); var io = require('socket.io').listen(server); var redis = require('redis'); var redisClient = redis.createClient(); io.sockets.on('connection', function(client) { redisClient.lrange('questions', 0, -1, function(err, questions){ questions.forEach(function(question){ client.emit("question", question); }); }); client.on('answer', function(question, answer) { client.broadcast.emit('answer', question, answer); }); client.on('question', function(question) { if(!client.question_asked) { client.question_asked = true; client.broadcast.emit('question', question); redisClient.lpush("questions", question); } }); });
Limiting Questions Stored
Great work! One last thing though, since every time a new question comes in we store it in the questions list, we might run into a problem where there are just too many questions stored in that list.
Add a callback to lpush
that will be used to limit the size of the list down to a max of 20.
Use the ltrim
command to limit the size of the list stored within redis
to a maximum size of 20.
redisClient.lpush("questions", question, function(){ redisClient.ltrim("questions", 0, 19); });
var express = require('express'); var app = express(); var server = require('http').createServer(app); var io = require('socket.io').listen(server); var redis = require('redis'); var redisClient = redis.createClient(); io.sockets.on('connection', function(client) { redisClient.lrange('questions', 0, -1, function(err, messages){ messages.forEach(function(m){ client.emit('question', m); }); }); client.on('answer', function(question, answer) { client.broadcast.emit('answer', question, answer); }); client.on('question', function(question) { if(!client.question_asked) { client.question_asked = true; client.broadcast.emit('question', question); redisClient.lpush("questions", question, function(){ redisClient.ltrim('questions', 0, 19); }); } }); });