mysqld_multi stop 1
followed by (repeatedly, a few moments later)
mysqld_multi report 1
returns:
Reporting MySQL servers
MySQL server from group: mysqld1 is running
The instances work fine and when I reboot, I have to start the instances to get them to come back up (a separate issue I need to resolve), but the stop
command doesn't seem to do anything.
It truly doesn't stop the instance because in the instance log file I see this:
120619 11:12:39 mysqld_safe A mysqld process already exists
after trying to run start
after.
Whenever this happens, I usually try to shutdown mysql using mysqladmin
For example, to shutdown mysql running on port 3307
Once you ran that make sure, mysql on that port is indeed down like this:
Give it a Try !!!
To stop any instance of
mysqld_multi
privileges should be granted tomulti_admin
user. Kindly check grants part also.This issue is still present in MYSQL 5.7.7 It sounds funny but the work around is to use pass instead of password in my.cnf for [multi_admin] group user.
after shutting down mysql service (using
mysqladmin -h127.0.0.1 -uroot -p shutdown
) and start it again, I could not connect to that server from external clientsI looked up the problem and I found this question, applied the answer, so that the problem is solved. that was because of waste DNS reverse query of mysqld.
mysqld_multi is just a perl script. Why
mysqld_multi start 1
andmysqld_multi report 1
working, butmysqld_multi stop 1
not, lies in my_print_defaults.In MySQL5.6+, password will be shown in "" by defaults,
mysqld_multi
start, report doesn't need authorization, so password is of no use, and it workers, but when stoping, user & password is needed for authorization, and "" breaks downmysqld_multi stop
.Solutions:
-s
option to my_print_defaults@RolandoMySQLDBA 's solution works, just because mysqld_multi will finally rely on mysqladmin to admin mysqld instance