A robust, performance-focused and full-featured Redis client for Node.js.
Supports Redis >= 2.6.12 and (Node.js >= 6). Completely compatible with Redis 6.x.
ioredis is a robust, full-featured Redis client that is used in the world's biggest online commerce company Alibaba and many other awesome companies.
- Full-featured. It supports Cluster, Sentinel, Streams, Pipelining and of course Lua scripting & Pub/Sub (with the support of binary messages).
- High performance.
- Delightful API. It works with Node callbacks and Native promises.
- Transformation of command arguments and replies.
- Transparent key prefixing.
- Abstraction for Lua scripting, allowing you to define custom commands.
- Support for binary data.
- Support for TLS 🔒.
- Support for offline queue and ready checking.
- Support for ES6 types, such as
Map
andSet
. - Support for GEO commands 📍.
- Support for Redis ACL.
- Sophisticated error handling strategy.
- Support for NAT mapping.
- Support for autopipelining
![Download on the App Store](resources/medis.png)
Looking for a Redis GUI manager for OS X, Windows and Linux? Here's Medis!
Medis is an open-sourced, beautiful, easy-to-use Redis GUI management application.
Medis starts with all the basic features you need:
- Keys viewing/editing
- SSH Tunnel for connecting with remote servers
- Terminal for executing custom commands
- JSON/MessagePack format viewing/editing and built-in highlighting/validator
- And other awesome features...
Medis is open sourced on GitHub
$ npm install ioredis
const Redis = require("ioredis");
const redis = new Redis(); // uses defaults unless given configuration object
// ioredis supports all Redis commands:
redis.set("foo", "bar"); // returns promise which resolves to string, "OK"
// the format is: redis[SOME_REDIS_COMMAND_IN_LOWERCASE](ARGUMENTS_ARE_JOINED_INTO_COMMAND_STRING)
// the js: ` redis.set("mykey", "Hello") ` is equivalent to the cli: ` redis> SET mykey "Hello" `
// ioredis supports the node.js callback style
redis.get("foo", function (err, result) {
if (err) {
console.error(err);
} else {
console.log(result); // Promise resolves to "bar"
}
});
// Or ioredis returns a promise if the last argument isn't a function
redis.get("foo").then(function (result) {
console.log(result); // Prints "bar"
});
// Most responses are strings, or arrays of strings
redis.zadd("sortedSet", 1, "one", 2, "dos", 4, "quatro", 3, "three");
redis.zrange("sortedSet", 0, 2, "WITHSCORES").then((res) => console.log(res)); // Promise resolves to ["one", "1", "dos", "2", "three", "3"] as if the command was ` redis> ZRANGE sortedSet 0 2 WITHSCORES `
// All arguments are passed directly to the redis server:
redis.set("key", 100, "EX", 10);
See the examples/
folder for more examples.
When a new Redis
instance is created,
a connection to Redis will be created at the same time.
You can specify which Redis to connect to by:
new Redis(); // Connect to 127.0.0.1:6379
new Redis(6380); // 127.0.0.1:6380
new Redis(6379, "192.168.1.1"); // 192.168.1.1:6379
new Redis("/tmp/redis.sock");
new Redis({
port: 6379, // Redis port
host: "127.0.0.1", // Redis host
family: 4, // 4 (IPv4) or 6 (IPv6)
password: "auth",
db: 0,
});
You can also specify connection options as a redis://
URL or rediss://
URL when using TLS encryption:
// Connect to 127.0.0.1:6380, db 4, using password "authpassword":
new Redis("redis://:[email protected]:6380/4");
// Username can also be passed via URI.
// It's worth to noticing that for compatibility reasons `allowUsernameInURI`
// need to be provided, otherwise the username part will be ignored.
new Redis(
"redis://username:[email protected]:6380/4?allowUsernameInURI=true"
);
See API Documentation for all available options.
Redis provides several commands for developers to implement the Publish–subscribe pattern. There are two roles in this pattern: publisher and subscriber. Publishers are not programmed to send their messages to specific subscribers. Rather, published messages are characterized into channels, without knowledge of what (if any) subscribers there may be.
By leveraging Node.js's built-in events module, ioredis makes pub/sub very straightforward to use. Below is a simple example that consists of two files, one is publisher.js that publishes messages to a channel, the other is subscriber.js that listens for messages on specific channels.
// publisher.js
const Redis = require("ioredis");
const redis = new Redis();
setInterval(() => {
const message = { foo: Math.random() };
// Publish to my-channel-1 or my-channel-2 randomly.
const channel = `my-channel-${1 + Math.round(Math.random())}`;
// Message can be either a string or a buffer
redis.publish(channel, JSON.stringify(message));
console.log("Published %s to %s", message, channel);
}, 1000);
// subscriber.js
const Redis = require("ioredis");
const redis = new Redis();
redis.subscribe("my-channel-1", "my-channel-2", (err, count) => {
if (err) {
// Just like other commands, subscribe() can fail for some reasons,
// ex network issues.
console.error("Failed to subscribe: %s", err.message);
} else {
// `count` represents the number of channels this client are currently subscribed to.
console.log(
`Subscribed successfully! This client is currently subscribed to ${count} channels.`
);
}
});
redis.on("message", (channel, message) => {
console.log(`Received ${message} from ${channel}`);
});
// There's also an event called 'messageBuffer', which is the same as 'message' except
// it returns buffers instead of strings.
// It's useful when the messages are binary data.
redis.on("messageBuffer", (channel, message) => {
// Both `channel` and `message` are buffers.
console.log(channel, message);
});
It worth noticing that a connection (aka Redis
instance) can't play both roles together. More specifically, when a client issues subscribe()
or psubscribe()
, it enters the "subscriber" mode. From that point, only commands that modify the subscription set are valid. Namely, they are: subscribe
, psubscribe
, unsubscribe
, punsubscribe
, ping
, and quit
. When the subscription set is empty (via unsubscribe
/punsubscribe
), the connection is put back into the regular mode.
If you want to do pub/sub in the same file/process, you should create a separate connection:
const Redis = require("ioredis");
const sub = new Redis();
const pub = new Redis();
sub.subscribe(/* ... */); // From now, `sub` enters the subscriber mode.
sub.on("message" /* ... */);
setInterval(() => {
// `pub` can be used to publish messages, or send other regular commands (e.g. `hgetall`)
// because it's not in the subscriber mode.
pub.publish(/* ... */);
}, 1000);
PSUBSCRIBE
is also supported in a similar way when you want to subscribe all channels whose name matches a pattern:
redis.psubscribe("pat?ern", (err, count) => {});
// Event names are "pmessage"/"pmessageBuffer" instead of "message/messageBuffer".
redis.on("pmessage", (pattern, channel, message) => {});
redis.on("pmessageBuffer", (pattern, channel, message) => {});
Redis v5 introduces a new data type called streams. It doubles as a communication channel for building streaming architectures and as a log-like data structure for persisting data. With ioredis, the usage can be pretty straightforward. Say we have a producer publishes messages to a stream with redis.xadd("mystream", "*", "randomValue", Math.random())
(You may find the official documentation of Streams as a starter to understand the parameters used), to consume the messages, we'll have a consumer with the following code:
const Redis = require("ioredis");
const redis = new Redis();
const processMessage = (message) => {
console.log("Id: %s. Data: %O", message[0], message[1]);
};
async function listenForMessage(lastId = "$") {
// `results` is an array, each element of which corresponds to a key.
// Because we only listen to one key (mystream) here, `results` only contains
// a single element. See more: https://redis.io/commands/xread#return-value
const results = await redis.xread("block", 0, "STREAMS", "mystream", lastId);
const [key, messages] = results[0]; // `key` equals to "mystream"
messages.forEach(processMessage);
// Pass the last id of the results to the next round.
await listenForMessage(messages[messages.length - 1][0]);
}
listenForMessage();
Arguments can be buffers:
redis.set("foo", Buffer.from("bar"));
And every command has a method that returns a Buffer (by adding a suffix of "Buffer" to the command name). To get a buffer instead of a utf8 string:
redis.getBuffer("foo", (err, result) => {
// result is a buffer.
});
If you want to send a batch of commands (e.g. > 5), you can use pipelining to queue the commands in memory and then send them to Redis all at once. This way the performance improves by 50%~300% (See benchmark section).
redis.pipeline()
creates a Pipeline
instance. You can call any Redis
commands on it just like the Redis
instance. The commands are queued in memory
and flushed to Redis by calling the exec
method:
const pipeline = redis.pipeline();
pipeline.set("foo", "bar");
pipeline.del("cc");
pipeline.exec((err, results) => {
// `err` is always null, and `results` is an array of responses
// corresponding to the sequence of queued commands.
// Each response follows the format `[err, result]`.
});
// You can even chain the commands:
redis
.pipeline()
.set("foo", "bar")
.del("cc")
.exec((err, results) => {});
// `exec` also returns a Promise:
const promise = redis.pipeline().set("foo", "bar").get("foo").exec();
promise.then((result) => {
// result === [[null, 'OK'], [null, 'bar']]
});
Each chained command can also have a callback, which will be invoked when the command gets a reply:
redis
.pipeline()
.set("foo", "bar")
.get("foo", (err, result) => {
// result === 'bar'
})
.exec((err, result) => {
// result[1][1] === 'bar'
});
In addition to adding commands to the pipeline
queue individually, you can also pass an array of commands and arguments to the constructor:
redis
.pipeline([
["set", "foo", "bar"],
["get", "foo"],
])
.exec(() => {
/* ... */
});
#length
property shows how many commands in the pipeline:
const length = redis.pipeline().set("foo", "bar").get("foo").length;
// length === 2
Most of the time, the transaction commands multi
& exec
are used together with pipeline.
Therefore, when multi
is called, a Pipeline
instance is created automatically by default,
so you can use multi
just like pipeline
:
redis
.multi()
.set("foo", "bar")
.get("foo")
.exec((err, results) => {
// results === [[null, 'OK'], [null, 'bar']]
});
If there's a syntax error in the transaction's command chain (e.g. wrong number of arguments, wrong command name, etc), then none of the commands would be executed, and an error is returned:
redis
.multi()
.set("foo")
.set("foo", "new value")
.exec((err, results) => {
// err:
// { [ReplyError: EXECABORT Transaction discarded because of previous errors.]
// name: 'ReplyError',
// message: 'EXECABORT Transaction discarded because of previous errors.',
// command: { name: 'exec', args: [] },
// previousErrors:
// [ { [ReplyError: ERR wrong number of arguments for 'set' command]
// name: 'ReplyError',
// message: 'ERR wrong number of arguments for \'set\' command',
// command: [Object] } ] }
});
In terms of the interface, multi
differs from pipeline
in that when specifying a callback
to each chained command, the queueing state is passed to the callback instead of the result of the command:
redis
.multi()
.set("foo", "bar", (err, result) => {
// result === 'QUEUED'
})
.exec(/* ... */);
If you want to use transaction without pipeline, pass { pipeline: false }
to multi
,
and every command will be sent to Redis immediately without waiting for an exec
invocation:
redis.multi({ pipeline: false });
redis.set("foo", "bar");
redis.get("foo");
redis.exec((err, result) => {
// result === [[null, 'OK'], [null, 'bar']]
});
The constructor of multi
also accepts a batch of commands:
redis
.multi([
["set", "foo", "bar"],
["get", "foo"],
])
.exec(() => {
/* ... */
});
Inline transactions are supported by pipeline, which means you can group a subset of commands in the pipeline into a transaction:
redis
.pipeline()
.get("foo")
.multi()
.set("foo", "bar")
.get("foo")
.exec()
.get("foo")
.exec();
ioredis supports all of the scripting commands such as EVAL
, EVALSHA
and SCRIPT
.
However, it's tedious to use in real world scenarios since developers have to take
care of script caching and to detect when to use EVAL
and when to use EVALSHA
.
ioredis exposes a defineCommand
method to make scripting much easier to use:
const redis = new Redis();
// This will define a command echo:
redis.defineCommand("echo", {
numberOfKeys: 2,
lua: "return {KEYS[1],KEYS[2],ARGV[1],ARGV[2]}",
});
// Now `echo` can be used just like any other ordinary command,
// and ioredis will try to use `EVALSHA` internally when possible for better performance.
redis.echo("k1", "k2", "a1", "a2", (err, result) => {
// result === ['k1', 'k2', 'a1', 'a2']
});
// `echoBuffer` is also defined automatically to return buffers instead of strings:
redis.echoBuffer("k1", "k2", "a1", "a2", (err, result) => {
// result[0] equals to Buffer.from('k1');
});
// And of course it works with pipeline:
redis.pipeline().set("foo", "bar").echo("k1", "k2", "a1", "a2").exec();
If the number of keys can't be determined when defining a command, you can
omit the numberOfKeys
property and pass the number of keys as the first argument
when you call the command:
redis.defineCommand("echoDynamicKeyNumber", {
lua: "return {KEYS[1],KEYS[2],ARGV[1],ARGV[2]}",
});
// Now you have to pass the number of keys as the first argument every time
// you invoke the `echoDynamicKeyNumber` command:
redis.echoDynamicKeyNumber(2, "k1", "k2", "a1", "a2", (err, result) => {
// result === ['k1', 'k2', 'a1', 'a2']
});
This feature allows you to specify a string that will automatically be prepended to all the keys in a command, which makes it easier to manage your key namespaces.
Warning This feature won't apply to commands like KEYS and SCAN that take patterns rather than actual keys(#239), and this feature also won't apply to the replies of commands even if they are key names (#325).
const fooRedis = new Redis({ keyPrefix: "foo:" });
fooRedis.set("bar", "baz"); // Actually sends SET foo:bar baz
fooRedis.defineCommand("echo", {
numberOfKeys: 2,
lua: "return {KEYS[1],KEYS[2],ARGV[1],ARGV[2]}",
});
// Works well with pipelining/transaction
fooRedis
.pipeline()
// Sends SORT foo:list BY foo:weight_*->fieldname
.sort("list", "BY", "weight_*->fieldname")
// Supports custom commands
// Sends EVALSHA xxx foo:k1 foo:k2 a1 a2
.echo("k1", "k2", "a1", "a2")
.exec();
Most Redis commands take one or more Strings as arguments,
and replies are sent back as a single String or an Array of Strings. However, sometimes
you may want something different. For instance, it would be more convenient if the HGETALL
command returns a hash (e.g. { key: val1, key2: v2 }
) rather than an array of key values (e.g. [key1, val1, key2, val2]
).
ioredis has a flexible system for transforming arguments and replies. There are two types of transformers, argument transformer and reply transformer:
const Redis = require("ioredis");
// Here's the built-in argument transformer converting
// hmset('key', { k1: 'v1', k2: 'v2' })
// or
// hmset('key', new Map([['k1', 'v1'], ['k2', 'v2']]))
// into
// hmset('key', 'k1', 'v1', 'k2', 'v2')
Redis.Command.setArgumentTransformer("hmset", (args) => {
if (args.length === 2) {
if (typeof Map !== "undefined" && args[1] instanceof Map) {
// utils is a internal module of ioredis
return [args[0]].concat(utils.convertMapToArray(args[1]));
}
if (typeof args[1] === "object" && args[1] !== null) {
return [args[0]].concat(utils.convertObjectToArray(args[1]));
}
}
return args;
});
// Here's the built-in reply transformer converting the HGETALL reply
// ['k1', 'v1', 'k2', 'v2']
// into
// { k1: 'v1', 'k2': 'v2' }
Redis.Command.setReplyTransformer("hgetall", (result) => {
if (Array.isArray(result)) {
const obj = {};
for (let i = 0; i < result.length; i += 2) {
obj[result[i]] = result[i + 1];
}
return obj;
}
return result;
});
There are three built-in transformers, two argument transformers for hmset
& mset
and
a reply transformer for hgetall
. Transformers for hmset
and hgetall
were mentioned
above, and the transformer for mset
is similar to the one for hmset
:
redis.mset({ k1: "v1", k2: "v2" });
redis.get("k1", (err, result) => {
// result === 'v1';
});
redis.mset(
new Map([
["k3", "v3"],
["k4", "v4"],
])
);
redis.get("k3", (err, result) => {
// result === 'v3';
});
Another useful example of a reply transformer is one that changes hgetall
to return array of arrays instead of objects which avoids a unwanted conversation of hash keys to strings when dealing with binary hash keys:
Redis.Command.setReplyTransformer("hgetall", (result) => {
const arr = [];
for (let i = 0; i < result.length; i += 2) {
arr.push([result[i], result[i + 1]]);
}
return arr;
});
redis.hset("h1", Buffer.from([0x01]), Buffer.from([0x02]));
redis.hset("h1", Buffer.from([0x03]), Buffer.from([0x04]));
redis.hgetallBuffer("h1", (err, result) => {
// result === [ [ <Buffer 01>, <Buffer 02> ], [ <Buffer 03>, <Buffer 04> ] ];
});
Redis supports the MONITOR command, which lets you see all commands received by the Redis server across all client connections, including from other client libraries and other computers.
The monitor
method returns a monitor instance.
After you send the MONITOR command, no other commands are valid on that connection. ioredis will emit a monitor event for every new monitor message that comes across.
The callback for the monitor event takes a timestamp from the Redis server and an array of command arguments.
Here is a simple example:
redis.monitor((err, monitor) => {
monitor.on("monitor", (time, args, source, database) => {});
});
Here is another example illustrating an async
function and monitor.disconnect()
:
async () => {
const monitor = await redis.monitor();
monitor.on("monitor", console.log);
// Any other tasks
monitor.disconnect();
};
Redis 2.8 added the SCAN
command to incrementally iterate through the keys in the database. It's different from KEYS
in that
SCAN
only returns a small number of elements each call, so it can be used in production without the downside
of blocking the server for a long time. However, it requires recording the cursor on the client side each time
the SCAN
command is called in order to iterate through all the keys correctly. Since it's a relatively common use case, ioredis
provides a streaming interface for the SCAN
command to make things much easier. A readable stream can be created by calling scanStream
:
const redis = new Redis();
// Create a readable stream (object mode)
const stream = redis.scanStream();
stream.on("data", (resultKeys) => {
// `resultKeys` is an array of strings representing key names.
// Note that resultKeys may contain 0 keys, and that it will sometimes
// contain duplicates due to SCAN's implementation in Redis.
for (let i = 0; i < resultKeys.length; i++) {
console.log(resultKeys[i]);
}
});
stream.on("end", () => {
console.log("all keys have been visited");
});
scanStream
accepts an option, with which you can specify the MATCH
pattern, the TYPE
filter, and the COUNT
argument:
const stream = redis.scanStream({
// only returns keys following the pattern of `user:*`
match: "user:*",
// only return objects that match a given type,
// (requires Redis >= 6.0)
type: "zset",
// returns approximately 100 elements per call
count: 100,
});
Just like other commands, scanStream
has a binary version scanBufferStream
, which returns an array of buffers. It's useful when
the key names are not utf8 strings.
There are also hscanStream
, zscanStream
and sscanStream
to iterate through elements in a hash, zset and set. The interface of each is
similar to scanStream
except the first argument is the key name:
const stream = redis.hscanStream("myhash", {
match: "age:??",
});
You can learn more from the Redis documentation.
Useful Tips
It's pretty common that doing an async task in the data
handler. We'd like the scanning process to be paused until the async task to be finished. Stream#pause()
and Stream#resume()
do the trick. For example if we want to migrate data in Redis to MySQL:
const stream = redis.scanStream();
stream.on("data", (resultKeys) => {
// Pause the stream from scanning more keys until we've migrated the current keys.
stream.pause();
Promise.all(resultKeys.map(migrateKeyToMySQL)).then(() => {
// Resume the stream here.
stream.resume();
});
});
stream.on("end", () => {
console.log("done migration");
});
By default, ioredis will try to reconnect when the connection to Redis is lost
except when the connection is closed manually by redis.disconnect()
or redis.quit()
.
It's very flexible to control how long to wait to reconnect after disconnection
using the retryStrategy
option:
const redis = new Redis({
// This is the default value of `retryStrategy`
retryStrategy(times) {
const delay = Math.min(times * 50, 2000);
return delay;
},
});
retryStrategy
is a function that will be called when the connection is lost.
The argument times
means this is the nth reconnection being made and
the return value represents how long (in ms) to wait to reconnect. When the
return value isn't a number, ioredis will stop trying to reconnect, and the connection
will be lost forever if the user doesn't call redis.connect()
manually.
When reconnected, the client will auto subscribe to channels that the previous connection subscribed to.
This behavior can be disabled by setting the autoResubscribe
option to false
.
And if the previous connection has some unfulfilled commands (most likely blocking commands such as brpop
and blpop
),
the client will resend them when reconnected. This behavior can be disabled by setting the autoResendUnfulfilledCommands
option to false
.
By default, all pending commands will be flushed with an error every 20 retry attempts. That makes sure commands won't wait forever when the connection is down. You can change this behavior by setting maxRetriesPerRequest
:
const redis = new Redis({
maxRetriesPerRequest: 1,
});
Set maxRetriesPerRequest to null
to disable this behavior, and every command will wait forever until the connection is alive again (which is the default behavior before ioredis v4).
Besides auto-reconnect when the connection is closed, ioredis supports reconnecting on certain Redis errors using the reconnectOnError
option. Here's an example that will reconnect when receiving READONLY
error:
const redis = new Redis({
reconnectOnError(err) {
const targetError = "READONLY";
if (err.message.includes(targetError)) {
// Only reconnect when the error contains "READONLY"
return true; // or `return 1;`
}
},
});
This feature is useful when using Amazon ElastiCache instances with Auto-failover disabled. On these instances, test your reconnectOnError
handler by manually promoting the replica node to the primary role using the AWS console. The following writes fail with the error READONLY
. Using reconnectOnError
, we can force the connection to reconnect on this error in order to connect to the new master. Furthermore, if the reconnectOnError
returns 2
, ioredis will resend the failed command after reconnecting.
On ElastiCache insances with Auto-failover enabled, reconnectOnError
does not execute. Instead of returning a Redis error, AWS closes all connections to the master endpoint until the new primary node is ready. ioredis reconnects via retryStrategy
instead of reconnectOnError
after about a minute. On ElastiCache insances with Auto-failover enabled, test failover events with the Failover primary
option in the AWS console.
The Redis instance will emit some events about the state of the connection to the Redis server.
Event | Description |
---|---|
connect | emits when a connection is established to the Redis server. |
ready | If enableReadyCheck is true , client will emit ready when the server reports that it is ready to receive commands (e.g. finish loading data from disk).Otherwise, ready will be emitted immediately right after the connect event. |
error | emits when an error occurs while connecting. However, ioredis emits all error events silently (only emits when there's at least one listener) so that your application won't crash if you're not listening to the error event. |
close | emits when an established Redis server connection has closed. |
reconnecting | emits after close when a reconnection will be made. The argument of the event is the time (in ms) before reconnecting. |
end | emits after close when no more reconnections will be made, or the connection is failed to establish. |
wait | emits when lazyConnect is set and will wait for the first command to be called before connecting. |
You can also check out the Redis#status
property to get the current connection status.
Besides the above connection events, there are several other custom events:
Event | Description |
---|---|
select | emits when the database changed. The argument is the new db number. |
When a command can't be processed by Redis (being sent before the ready
event), by default, it's added to the offline queue and will be
executed when it can be processed. You can disable this feature by setting the enableOfflineQueue
option to false
:
const redis = new Redis({ enableOfflineQueue: false });
Redis doesn't support TLS natively, however if the redis server you want to connect to is hosted behind a TLS proxy (e.g. stunnel) or is offered by a PaaS service that supports TLS connection (e.g. Redis Labs), you can set the tls
option:
const redis = new Redis({
host: "localhost",
tls: {
// Refer to `tls.connect()` section in
// https://nodejs.org/api/tls.html
// for all supported options
ca: fs.readFileSync("cert.pem"),
},
});
Alternatively, specify the connection through a rediss://
URL.
const redis = new Redis("rediss://redis.my-service.com");
ioredis supports Sentinel out of the box. It works transparently as all features that work when you connect to a single node also work when you connect to a sentinel group. Make sure to run Redis >= 2.8.12 if you want to use this feature. Sentinels have a default port of 26379.
To connect using Sentinel, use:
const redis = new Redis({
sentinels: [
{ host: "localhost", port: 26379 },
{ host: "localhost", port: 26380 },
],
name: "mymaster",
});
redis.set("foo", "bar");
The arguments passed to the constructor are different from the ones you use to connect to a single node, where:
name
identifies a group of Redis instances composed of a master and one or more slaves (mymaster
in the example);sentinelPassword
(optional) password for Sentinel instances.sentinels
are a list of sentinels to connect to. The list does not need to enumerate all your sentinel instances, but a few so that if one is down the client will try the next one.role
(optional) with a value ofslave
will return a random slave from the Sentinel group.preferredSlaves
(optional) can be used to prefer a particular slave or set of slaves based on priority. It accepts a function or array.
ioredis guarantees that the node you connected to is always a master even after a failover. When a failover happens, instead of trying to reconnect to the failed node (which will be demoted to slave when it's available again), ioredis will ask sentinels for the new master node and connect to it. All commands sent during the failover are queued and will be executed when the new connection is established so that none of the commands will be lost.
It's possible to connect to a slave instead of a master by specifying the option role
with the value of slave
and ioredis will try to connect to a random slave of the specified master, with the guarantee that the connected node is always a slave. If the current node is promoted to master due to a failover, ioredis will disconnect from it and ask the sentinels for another slave node to connect to.
If you specify the option preferredSlaves
along with role: 'slave'
ioredis will attempt to use this value when selecting the slave from the pool of available slaves. The value of preferredSlaves
should either be a function that accepts an array of available slaves and returns a single result, or an array of slave values priorities by the lowest prio
value first with a default value of 1
.
// available slaves format
const availableSlaves = [{ ip: "127.0.0.1", port: "31231", flags: "slave" }];
// preferredSlaves array format
let preferredSlaves = [
{ ip: "127.0.0.1", port: "31231", prio: 1 },
{ ip: "127.0.0.1", port: "31232", prio: 2 },
];
// preferredSlaves function format
preferredSlaves = function (availableSlaves) {
for (let i = 0; i < availableSlaves.length; i++) {
const slave = availableSlaves[i];
if (slave.ip === "127.0.0.1") {
if (slave.port === "31234") {
return slave;
}
}
}
// if no preferred slaves are available a random one is used
return false;
};
const redis = new Redis({
sentinels: [
{ host: "127.0.0.1", port: 26379 },
{ host: "127.0.0.1", port: 26380 },
],
name: "mymaster",
role: "slave",
preferredSlaves: preferredSlaves,
});
Besides the retryStrategy
option, there's also a sentinelRetryStrategy
in Sentinel mode which will be invoked when all the sentinel nodes are unreachable during connecting. If sentinelRetryStrategy
returns a valid delay time, ioredis will try to reconnect from scratch. The default value of sentinelRetryStrategy
is:
function (times) {
const delay = Math.min(times * 10, 1000);
return delay;
}
Redis Cluster provides a way to run a Redis installation where data is automatically sharded across multiple Redis nodes. You can connect to a Redis Cluster like this:
const Redis = require("ioredis");
const cluster = new Redis.Cluster([
{
port: 6380,
host: "127.0.0.1",
},
{
port: 6381,
host: "127.0.0.1",
},
]);
cluster.set("foo", "bar");
cluster.get("foo", (err, res) => {
// res === 'bar'
});
Cluster
constructor accepts two arguments, where:
-
The first argument is a list of nodes of the cluster you want to connect to. Just like Sentinel, the list does not need to enumerate all your cluster nodes, but a few so that if one is unreachable the client will try the next one, and the client will discover other nodes automatically when at least one node is connected.
-
The second argument is the options, where:
-
clusterRetryStrategy
: When none of the startup nodes are reachable,clusterRetryStrategy
will be invoked. When a number is returned, ioredis will try to reconnect to the startup nodes from scratch after the specified delay (in ms). Otherwise, an error of "None of startup nodes is available" will be returned. The default value of this option is:function (times) { const delay = Math.min(100 + times * 2, 2000); return delay; }
It's possible to modify the
startupNodes
property in order to switch to another set of nodes here:function (times) { this.startupNodes = [{ port: 6790, host: '127.0.0.1' }]; return Math.min(100 + times * 2, 2000); }
-
dnsLookup
: Alternative DNS lookup function (dns.lookup()
is used by default). It may be useful to override this in special cases, such as when AWS ElastiCache used with TLS enabled. -
enableOfflineQueue
: Similar to theenableOfflineQueue
option ofRedis
class. -
enableReadyCheck
: When enabled, "ready" event will only be emitted whenCLUSTER INFO
command reporting the cluster is ready for handling commands. Otherwise, it will be emitted immediately after "connect" is emitted. -
scaleReads
: Config where to send the read queries. See below for more details. -
maxRedirections
: When a cluster related error (e.g.MOVED
,ASK
andCLUSTERDOWN
etc.) is received, the client will redirect the command to another node. This option limits the max redirections allowed when sending a command. The default value is16
. -
retryDelayOnFailover
: If the target node is disconnected when sending a command, ioredis will retry after the specified delay. The default value is100
. You should make sureretryDelayOnFailover * maxRedirections > cluster-node-timeout
to insure that no command will fail during a failover. -
retryDelayOnClusterDown
: When a cluster is down, all commands will be rejected with the error ofCLUSTERDOWN
. If this option is a number (by default, it is100
), the client will resend the commands after the specified time (in ms). -
retryDelayOnTryAgain
: If this option is a number (by default, it is100
), the client will resend the commands rejected withTRYAGAIN
error after the specified time (in ms). -
retryDelayOnMoved
: By default, this value is0
(in ms), which means when aMOVED
error is received, the client will resend the command instantly to the node returned together with theMOVED
error. However, sometimes it takes time for a cluster to become state stabilized after a failover, so adding a delay before resending can prevent a ping pong effect. -
redisOptions
: Default options passed to the constructor ofRedis
when connecting to a node. -
slotsRefreshTimeout
: Milliseconds before a timeout occurs while refreshing slots from the cluster (default1000
). -
slotsRefreshInterval
: Milliseconds between every automatic slots refresh (default5000
).
-
A typical redis cluster contains three or more masters and several slaves for each master. It's possible to scale out redis cluster by sending read queries to slaves and write queries to masters by setting the scaleReads
option.
scaleReads
is "master" by default, which means ioredis will never send any queries to slaves. There are other three available options:
- "all": Send write queries to masters and read queries to masters or slaves randomly.
- "slave": Send write queries to masters and read queries to slaves.
- a custom
function(nodes, command): node
: Will choose the custom function to select to which node to send read queries (write queries keep being sent to master). The first node innodes
is always the master serving the relevant slots. If the function returns an array of nodes, a random node of that list will be selected.
For example:
const cluster = new Redis.Cluster(
[
/* nodes */
],
{
scaleReads: "slave",
}
);
cluster.set("foo", "bar"); // This query will be sent to one of the masters.
cluster.get("foo", (err, res) => {
// This query will be sent to one of the slaves.
});
NB In the code snippet above, the res
may not be equal to "bar" because of the lag of replication between the master and slaves.
Every command will be sent to exactly one node. For commands containing keys, (e.g. GET
, SET
and HGETALL
), ioredis sends them to the node that serving the keys, and for other commands not containing keys, (e.g. INFO
, KEYS
and FLUSHDB
), ioredis sends them to a random node.
Sometimes you may want to send a command to multiple nodes (masters or slaves) of the cluster, you can get the nodes via Cluster#nodes()
method.
Cluster#nodes()
accepts a parameter role, which can be "master", "slave" and "all" (default), and returns an array of Redis
instance. For example:
// Send `FLUSHDB` command to all slaves:
const slaves = cluster.nodes("slave");
Promise.all(slaves.map((node) => node.flushdb()));
// Get keys of all the masters:
const masters = cluster.nodes("master");
Promise.all(
masters
.map((node) => node.keys())
.then((keys) => {
// keys: [['key1', 'key2'], ['key3', 'key4']]
})
);
Sometimes the cluster is hosted within a internal network that can only be accessed via a NAT (Network Address Translation) instance. See Accessing ElastiCache from outside AWS as an example.
You can specify nat mapping rules via natMap
option:
const cluster = new Redis.Cluster(
[
{
host: "203.0.113.73",
port: 30001,
},
],
{
natMap: {
"10.0.1.230:30001": { host: "203.0.113.73", port: 30001 },
"10.0.1.231:30001": { host: "203.0.113.73", port: 30002 },
"10.0.1.232:30001": { host: "203.0.113.73", port: 30003 },
},
}
);
This option is also useful when the cluster is running inside a Docker container.
Almost all features that are supported by Redis
are also supported by Redis.Cluster
, e.g. custom commands, transaction and pipeline.
However there are some differences when using transaction and pipeline in Cluster mode:
- All keys in a pipeline should belong to slots served by the same node, since ioredis sends all commands in a pipeline to the same node.
- You can't use
multi
without pipeline (akacluster.multi({ pipeline: false })
). This is because when you callcluster.multi({ pipeline: false })
, ioredis doesn't know which node themulti
command should be sent to.
When any commands in a pipeline receives a MOVED
or ASK
error, ioredis will resend the whole pipeline to the specified node automatically if all of the following conditions are satisfied:
- All errors received in the pipeline are the same. For example, we won't resend the pipeline if we got two
MOVED
errors pointing to different nodes. - All commands executed successfully are readonly commands. This makes sure that resending the pipeline won't have side effects.
Pub/Sub in cluster mode works exactly as the same as in standalone mode. Internally, when a node of the cluster receives a message, it will broadcast the message to the other nodes. ioredis makes sure that each message will only be received once by strictly subscribing one node at the same time.
const nodes = [
/* nodes */
];
const pub = new Redis.Cluster(nodes);
const sub = new Redis.Cluster(nodes);
sub.on("message", (channel, message) => {
console.log(channel, message);
});
sub.subscribe("news", () => {
pub.publish("news", "highlights");
});
Event | Description |
---|---|
connect | emits when a connection is established to the Redis server. |
ready | emits when CLUSTER INFO reporting the cluster is able to receive commands (if enableReadyCheck is true ) or immediately after connect event (if enableReadyCheck is false). |
error | emits when an error occurs while connecting with a property of lastNodeError representing the last node error received. This event is emitted silently (only emitting if there's at least one listener). |
close | emits when an established Redis server connection has closed. |
reconnecting | emits after close when a reconnection will be made. The argument of the event is the time (in ms) before reconnecting. |
end | emits after close when no more reconnections will be made. |
+node | emits when a new node is connected. |
-node | emits when a node is disconnected. |
node error | emits when an error occurs when connecting to a node. The second argument indicates the address of the node. |
Setting the password
option to access password-protected clusters:
const Redis = require("ioredis");
const cluster = new Redis.Cluster(nodes, {
redisOptions: {
password: "your-cluster-password",
},
});
If some of nodes in the cluster using a different password, you should specify them in the first parameter:
const Redis = require("ioredis");
const cluster = new Redis.Cluster(
[
// Use password "password-for-30001" for 30001
{ port: 30001, password: "password-for-30001" },
// Don't use password when accessing 30002
{ port: 30002, password: null },
// Other nodes will use "fallback-password"
],
{
redisOptions: {
password: "fallback-password",
},
}
);
AWS ElastiCache for Redis (Clustered Mode) supports TLS encryption. If you use
this, you may encounter errors with invalid certificates. To resolve this
issue, construct the Cluster
with the dnsLookup
option as follows:
const cluster = new Redis.Cluster(
[
{
host: "clustercfg.myCluster.abcdefg.xyz.cache.amazonaws.com",
port: 6379,
},
],
{
dnsLookup: (address, callback) => callback(null, address),
redisOptions: {
tls: {},
},
}
);
In standard mode, when you issue multiple commands, ioredis sends them to the server one by one. As described in Redis pipeline documentation, this is a suboptimal use of the network link, especially when such link is not very performant.
The TCP and network overhead negatively affects performance. Commands are stuck in the send queue until the previous ones are correctly delivered to the server. This is a problem known as Head-Of-Line blocking (HOL).
ioredis supports a feature called “auto pipelining”. It can be enabled by setting the option enableAutoPipelining
to true
. No other code change is necessary.
In auto pipelining mode, all commands issued during an event loop are enqueued in a pipeline automatically managed by ioredis. At the end of the iteration, the pipeline is executed and thus all commands are sent to the server at the same time.
This feature can dramatically improve throughput and avoids HOL blocking. In our benchmarks, the improvement was between 35% and 50%.
While an automatic pipeline is executing, all new commands will be enqueued in a new pipeline which will be executed as soon as the previous finishes.
When using Redis Cluster, one pipeline per node is created. Commands are assigned to pipelines according to which node serves the slot.
A pipeline will thus contain commands using different slots but that ultimately are assigned to the same node.
Note that the same slot limitation within a single command still holds, as it is a Redis limitation.
This sample code uses ioredis with automatic pipeline enabled.
const Redis = require("./built");
const http = require("http");
const db = new Redis({ enableAutoPipelining: true });
const server = http.createServer((request, response) => {
const key = new URL(request.url, "https://localhost:3000/").searchParams.get(
"key"
);
db.get(key, (err, value) => {
response.writeHead(200, { "Content-Type": "text/plain" });
response.end(value);
});
});
server.listen(3000);
When Node receives requests, it schedules them to be processed in one or more iterations of the events loop.
All commands issued by requests processing during one iteration of the loop will be wrapped in a pipeline automatically created by ioredis.
In the example above, the pipeline will have the following contents:
GET key1
GET key2
GET key3
...
GET keyN
When all events in the current loop have been processed, the pipeline is executed and thus all commands are sent to the server at the same time.
While waiting for pipeline response from Redis, Node will still be able to process requests. All commands issued by request handler will be enqueued in a new automatically created pipeline. This pipeline will not be sent to the server yet.
As soon as a previous automatic pipeline has received all responses from the server, the new pipeline is immediately sent without waiting for the events loop iteration to finish.
This approach increases the utilization of the network link, reduces the TCP overhead and idle times and therefore improves throughput.
Here's some of the results of our tests for a single node.
Each iteration of the test runs 1000 random commands on the server.
Samples | Result | Tolerance | |
---|---|---|---|
default | 1000 | 174.62 op/sec | ± 0.45 % |
enableAutoPipelining=true | 1500 | 233.33 op/sec | ± 0.88 % |
And here's the same test for a cluster of 3 masters and 3 replicas:
Samples | Result | Tolerance | |
---|---|---|---|
default | 1000 | 164.05 op/sec | ± 0.42 % |
enableAutoPipelining=true | 3000 | 235.31 op/sec | ± 0.94 % |
All the errors returned by the Redis server are instances of ReplyError
, which can be accessed via Redis
:
const Redis = require("ioredis");
const redis = new Redis();
// This command causes a reply error since the SET command requires two arguments.
redis.set("foo", (err) => {
err instanceof Redis.ReplyError;
});
This is the error stack of the ReplyError
:
ReplyError: ERR wrong number of arguments for 'set' command
at ReplyParser._parseResult (/app/node_modules/ioredis/lib/parsers/javascript.js:60:14)
at ReplyParser.execute (/app/node_modules/ioredis/lib/parsers/javascript.js:178:20)
at Socket.<anonymous> (/app/node_modules/ioredis/lib/redis/event_handler.js:99:22)
at Socket.emit (events.js:97:17)
at readableAddChunk (_stream_readable.js:143:16)
at Socket.Readable.push (_stream_readable.js:106:10)
at TCP.onread (net.js:509:20)
By default, the error stack doesn't make any sense because the whole stack happens in the ioredis
module itself, not in your code. So it's not easy to find out where the error happens in your code.
ioredis provides an option showFriendlyErrorStack
to solve the problem. When you enable
showFriendlyErrorStack
, ioredis will optimize the error stack for you:
const Redis = require("ioredis");
const redis = new Redis({ showFriendlyErrorStack: true });
redis.set("foo");
And the output will be:
ReplyError: ERR wrong number of arguments for 'set' command
at Object.<anonymous> (/app/index.js:3:7)
at Module._compile (module.js:446:26)
at Object.Module._extensions..js (module.js:464:10)
at Module.load (module.js:341:32)
at Function.Module._load (module.js:296:12)
at Function.Module.runMain (module.js:487:10)
at startup (node.js:111:16)
at node.js:799:3
This time the stack tells you that the error happens on the third line in your code. Pretty sweet! However, it would decrease the performance significantly to optimize the error stack. So by default, this option is disabled and can only be used for debugging purposes. You shouldn't use this feature in a production environment.
If you're an advanced user, you may want to plug in your own promise library like bluebird. Just set Redis.Promise to your favorite ES6-style promise constructor and ioredis will use it.
const Redis = require("ioredis");
Redis.Promise = require("bluebird");
const redis = new Redis();
// Use bluebird
assert.equal(redis.get().constructor, require("bluebird"));
// You can change the Promise implementation at any time:
Redis.Promise = global.Promise;
assert.equal(redis.get().constructor, global.Promise);
Start a Redis server on 127.0.0.1:6379, and then:
$ npm test
FLUSH ALL
will be invoked after each test, so make sure there's no valuable data in it before running tests.
If your testing environment does not let you spin up a Redis server ioredis-mock is a drop-in replacement you can use in your tests. It aims to behave identically to ioredis connected to a Redis server so that your integration tests is easier to write and of better quality.
You can set the DEBUG
env to ioredis:*
to print debug info:
$ DEBUG=ioredis:* node app.js
I'm happy to receive bug reports, fixes, documentation enhancements, and any other improvements.
And since I'm not a native English speaker, if you find any grammar mistakes in the documentation, please also let me know. :)
Open source is hard and time-consuming. If you want to invest in ioredis's future you can become a sponsor and make us spend more time on this library's improvements and new features.
Thank you for using ioredis :-)
This project exists thanks to all the people who contribute:
MIT