Allow wsrelay to fail without FATAL (#15191)

We have not identify the root cause of wsrelay failure but attempt to make wsrelay restart itself resulted in postgres and redis connection leak. We were not able to fully identify where the redis connection leak comes from so reverting back to failing and removing startsecs 30 will prevent wsrelay to FATAL

tools/ansible/roles/dockerfile/templates/supervisor_task.conf.j2

@@ -31,7 +31,6 @@ command = awx-manage run_wsrelay
 directory = /var/lib/awx
 {% endif %}
 autorestart = true
-startsecs = 30
 stopasgroup=true
 killasgroup=true
 stdout_logfile=/dev/stdout