Try to gracefully shut the process down/ start it back up by logging into the command line as user root, run service salWatchdog restart. If the service will not stop, run ps -ef |grep java to locate the process ID (PID).
If you do not see a salWatchdog PID, navigate to /opt/avaya/SAL/gateway/SALWatchdog/pids and cat these two files; AgentGateway.pid and SALWatchdog.java.pid to obtain the process ID(s). Attempt to gracefully shut down the salWatchdog again by running kill -15 (PID). Status the service to ensure it is shut down, service salWatchdog status. If the service is still running, kill -9 (PID) as a last resort, status once more.
Null out the two .pid files (place a > before the file name, hit Enter, cat the file again to ensure it is now empty). Ensure the files are both owned by saluser:salgroup as the example below outlines, chown and chmod accordingly, chown saluser:salgroup AgentGateway.pid (failure to do so correctly could result in a "cannot read /opt/avaya/SAL/gateway/SALWatchdog/pids/AgentGateway.pid" error)
[root@SAL pids]# ll
total 8
-rw------- 1 saluser salgroup 6 Jun 17 11:45 AgentGateway.pid
-rw------- 1 saluser salgroup 6 Jun 17 11:45 SALWatchdog.java.pid
With the salWatchdog service now stopped, the PIDs removed, and the ownership/ permissions of the .pid files verified/ corrected, you should now be able to start the salWatchdog service, service salWatchdog start.