replicate.sh won't stop primary controller properly if it's ran via controller.sh

Ran into a bug that prevents proper setup of DB slave on primary node. Steps to replicate

1. Primary controller installed with platform-admin
2. Install init services on primary node (`sudo ./install-init.sh -s`)
2. Run replicate.sh for the first time (`./replicate.sh -s node2 -e http://:`)
3. install_init.sh on the secondary node (`sudo ./install-init.sh -s`)
4. Run replicate.sh in "final" mode (`./replicate.sh -s node2 -e http://:`)

Results in:
```
-- setting up primary slave
ERROR 1794 (HY000) at line 1: Slave is not configured or failed to initialize properly. You must at least set --server-id to enable either a master or a slave. Additional error messages can be found in the MySQL error log.
-- copy controller.sh to secondary
-- disable secondary appserver
-- remove secondary master.info
-- start secondary database
Connection to node2 closed.
-- wait for secondary to start
-- setting up secondary slave
-- removing skip-slave-start from primary
-- removing skip-slave-start from secondary
-- start primary slave
ERROR 1794 (HY000) at line 1: Slave is not configured or failed to initialize properly. You must at least set --server-id to enable either a master or a slave. Additional error messages can be found in the MySQL error log.
-- start secondary slave
-- primary slave status
-- secondary slave status
Slave_IO_State: Connecting to master
Master_Host: node1
Seconds_Behind_Master: 0
Master_Server_Id: 0
```

What happens is - "stopping primary" step fails as script:

1. in `stop_appdynamics_services`, tries to run `service appdcontroller stop` and `service appdcontroller-db stop` which both return 0 code, but do nothing - controller wasn't started by them
2. when the `stop_appdynamics_services` function returns, it returns 0 - meaning that `stop_appdynamics_services || $APPD_ROOT/bin/controller.sh stop` doesn't enter the second branch, so the controller and DB are still running during replication, but what is more - DB is not bounced with new db.cnf, hence slave setup error.