Local First Distribute Lock

Local first distribute lock aimed to reduce distribute-lock-server's concurrency pressure by using local lock first.As illustrated in figure 1, when a request want to fetch distribute lock, this request must first fetch JVM lock(local lock).By doing this, for a specific lock key, the distribute-lock-server will only bear a fixed number of concurrent machines, rather than requesting concurrent requests.

figure 1

This project was not started from scratch, but combined with existing distributed locks.Specifically, as follows.

Combine with zookeeper

This combination provides the following capabilities:

readWriteLock

lock

Combine with redisson

This combination provides the following capabilities:

readWriteLock

lock

spinLock

Combine with db

This combination provides the following capabilities:

spinLock