Fix health probe for several conductor workers

It was observed that when increasing amount of
conductor workers from default "1" to higher value
the readiness probe fails to check rabbitmq connections
for conductor processes - it happens since the script is trying
to obtain rabbitmq connections for parent conductor process
which in case of workers>1 doesn`t open rabbit connections
but spawns child processes which handle rabbitmq
connections instead.
This patch removes the "check-all-pids" option, keeps the logic
but simplifies and fastens he code - instead of checking all
processes when "check-all-pids" option was set (however
regardless of "sock_count value" if only one process opens connection
the check returns positive result) processes will be checked one-by-one
until the first one with open rabbitmq connection(s) is
found.

Change-Id: I72be0bbdefcba77a55b6ceed6e192c9621c069eb

nova/templates/bin/_health-probe.py.tpl

@@ -97,18 +97,10 @@ def check_service_status(transport):
 
 def tcp_socket_status(process, ports):
     """Check the tcp socket status on a process"""
-    sock_count = 0
-    parentId = 0
     for p in psutil.process_iter():
         try:
             with p.oneshot():
                 if process in " ".join(p.cmdline()):
-                    if parentId == 0:
-                        parentId = p.pid
-                    else:
-                        if p.ppid() == parentId and \
-                                not cfg.CONF.check_all_pids:
-                            continue
                     pcon = p.connections()
                     for con in pcon:
                         try:
@@ -117,14 +109,10 @@ def tcp_socket_status(process, ports):
                         except IndexError:
                             continue
                         if rport in ports and status == tcp_established:
-                            sock_count = sock_count + 1
+                            return 1
         except psutil.Error:
             continue
-
-    if sock_count == 0:
-        return 0
-    else:
-        return 1
+    return 0
 
 
 def configured_port_in_conf():
@@ -198,8 +186,6 @@ def test_rpc_liveness():
     cfg.CONF.register_cli_opt(cfg.StrOpt('service-queue-name'))
     cfg.CONF.register_cli_opt(cfg.BoolOpt('liveness-probe', default=False,
                                           required=False))
-    cfg.CONF.register_cli_opt(cfg.BoolOpt('check-all-pids', default=False,
-                                          required=False))
     cfg.CONF.register_cli_opt(cfg.BoolOpt('use-fqdn', default=False,
                                           required=False))
 

nova/templates/deployment-scheduler.yaml

@@ -68,7 +68,6 @@ spec:
                 - /etc/nova/nova.conf
                 - --service-queue-name
                 - scheduler
-                - --check-all-pids
             initialDelaySeconds: 80
             periodSeconds: 90
             timeoutSeconds: 70
@@ -82,7 +81,6 @@ spec:
                 - --service-queue-name
                 - scheduler
                 - --liveness-probe
-                - --check-all-pids
             initialDelaySeconds: 120
             periodSeconds: 90
             timeoutSeconds: 70