Fix port health check false negatives

Most of the opened port health checks are failing because containers
are not run under root user. This patch makes it to avoid use of lsof
so that all containers are checked correctly on opened ports.

Closes-bug: #1714077
Co-Authored-By: Derek Higgins <derekh@redhat.com>
Change-Id: I01389ce87f81486fd887d71e6816df76276011e0

healthcheck/common.sh

@@ -12,13 +12,17 @@ healthcheck_curl () {
 
 healthcheck_port () {
   process=$1
-  shift 1
 
-  port_args=()
-  for arg in "$@"; do
-    port_args+=("-i" "tcp:${arg}")
-  done
-  lsof +c0 -nP "${port_args[@]}" | awk '{print $1}' | grep -q "^${process}$"
+  # ss truncate command name to 15 characters and this behaviour
+  # cannot be diabled
+  if [ ${#process} -gt 15 ] ; then
+    process=${process:0:15}
+  fi
+
+  shift 1
+  args=$@
+  ports=${args// /|}
+  ss -ntp | awk '{print $5,"-",$6}' | egrep ":($ports)" | grep "$process"
 }
 
 get_config_val () {