High Availability Failure

[ocdkerosine]

Redlock is instantiated with three independent Redis instances. They run on different containers on different ports. When i shutdown one of the instances and try to acquire a lock it simply throws out EAI_AGAIN errors until i restart the instance and then the lock is successfully acquired or sometimes not.

If it is supposed to failover why does this behaviour exist? Instantiation below

        LockService.lock = new Redlock(
            [LockService.redis, LockService.redis2, LockService.redis3],
            {
                driftFactor: LockService.driftFactor, //all default values from repo
                retryCount: LockService.retryCount,
                retryDelay: LockService.retryDelay,
                retryJitter: LockService.retryJitter,
                automaticExtensionThreshold:
                    LockService.automaticExtensionThreshold,
            }
        );

After Shutting down one instance:

After restarting instance: