Battle-hardened distributed locking using redis
npm install node-redis-warlockwarlock
=======
Battle-hardened distributed locking using redis.
* node-redis compatible with v0.10
* Redis v2.6.12 or above. If you're running a Redis version from v2.6.0 to v2.6.11 inclusive use v0.0.7 of this module.
npm install node-redis-warlock
``javascript
const Warlock = require('node-redis-warlock');
const Redis = require('redis');
// Establish a redis client and pass it to warlock
const redis = Redis.createClient();
const warlock = Warlock(redis);
// Set a lock
const key = 'test-lock';
const ttl = 10000; // Lifetime of the lock
warlock.lock(key, ttl, (err, unlock) => {
if (err) {
// Something went wrong and we weren't able to set a lock
return;
}
if (typeof unlock === 'function') {
// If the lock is set successfully by this process, an unlock function is passed to our callback.
// Do the work that required lock protection, and then unlock() when finished...
//
// do stuff...
//
unlock();
} else {
// Otherwise, the lock was not established by us so we must decide what to do
// Perhaps wait a bit & retry...
}
});
// set a lock optimistically
const key = 'opt-lock';
const ttl = 10000;
const maxAttempts = 4; // Max number of times to try setting the lock before erroring
const wait = 1000; // Time to wait before another attempt if lock already in place
warlock.optimistic(key, ttl, maxAttempts, wait, (err, unlock) => {});
// unlock using the lock id
var key = 'test-lock-2';
var ttl = 10000;
let lockId;
warlock.lock(key, ttl, (err, _, id) => {
lockId = id;
});
// each client who knows the lockId can release the lock
warlock.unlock(key, lockId, (err, result) => {
if (result == 1) {
// unlocked successfully
}
});
// change a lock's ttl
var key = 'touch-lock';
var ttl = 10000;
var ttl2 = 20000;
warlock.lock(key, ttl, function(err, unlock, id) {
warlock.touch(key, id, ttl2, function(err) {});
});
``
* Read my Distributed locks using Redis article and Redis' author's A proposal for more reliable locks using Redis to learn more about the theory of distributed locks using Redis.