How to implement distributed locking in redis

Redis uses the SETNX command to set a key value in the absence of the existence of the SETNX command to implement a distributed lock: the SETNX command tries to acquire the lock, and if it is successful, the client will obtain the lock; set an expiration time for the lock to prevent deadlock; periodically renew the expiration time of the lock to ensure that the client holds the lock; when releasing the lock, delete the lock's key.

How to implement distributed locks in Redis

Distributed lock is a synchronization mechanism that ensures that in a distributed system, only one client can obtain access to shared resources at the same time. Redis implements distributed locking through the following mechanism:

SETNX Command

The SETNX command is an atomic operation that sets a key value in the absence of it. If the key already exists, the command will fail. Using the SETNX command, the client can try to acquire a distributed lock, as shown below:

 <code>SETNX my_lock 1</code>

If the my_lock key does not exist, the command will succeed and the client will obtain the lock. Otherwise, the command will fail, indicating that the lock has been held by other clients.

Expiration time

To prevent deadlocks, distributed locks are usually set to have an expiration time. When the client acquires the lock, it sets an expiration time for the lock so that the lock will be automatically released after that time.

 <code>SETEX my_lock 300 1</code>

This command will set a 5-minute expiration time for my_lock. If the client does not release the lock within the expiration time, Redis will automatically release the lock so that it can be used by other clients.

Polling and renewal

To ensure that the client still holds the lock after the lock expires, the client needs to periodically renew the expiration time of the lock. This can be done by doing the following:

 <code>WHILE my_lock == 1 SETEX my_lock 300 1 END</code>

The code will continue to renew the expiration time of my_lock as long as it is equal to the value set by the current client (i.e. 1).

Release the lock

When the client no longer needs the lock, it should release the lock by:

 <code>DEL my_lock</code>

This will remove the my_lock key, release the lock and make it available to other clients.

The above is the detailed content of How to implement distributed locking in redis. For more information, please follow other related articles on the PHP Chinese website!