redis fault handling process is already running or crashed

  Redis of a test library online today, after adding auth password authentication, I want to restart redis, it will not take effect without restarting. The result is (error) NOAUTH Authentication required no matter how you do it. redis-cli -a 123456 shutdown turned out to be useless. In a fit of rage, it was forced to close, kill -9 redis pid number.

As a result, it was a dog, and it was time to start. Consistent error /var/run/redis_6379.pid exists, process is already running or crashed.

Since it is a test library, the data in it is not too important, and the solution is relatively crude.

I directly back up the dump.rdb file to another directory,

then delete (rm -rf /var/run/redis_6379.pid) the pid number of redis and start

directly [root@localhost 6379]# /etc/init.d /redis_6379 start

starts normally.