Bug 1725900

Description of problem:
healthcheck for nova-api-metadata checks with `pgrep -f nova-metadata` but should pgrep for `nova-api-metadata`

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:

Upstream, this looks like this:
https://github.com/openstack/tripleo-common/blob/master/healthcheck/nova-metadata
~~~
#!/bin/bash

. ${HEALTHCHECK_SCRIPTS:-/usr/share/openstack-tripleo-common/healthcheck}/common.sh

check_url=$(get_url_from_vhost /etc/httpd/conf.d/10-nova_metadata_wsgi.conf)
healthcheck_curl ${check_url}
~~~

In OSP 13 and 14, these seem to run from the same image:
~~~
bad44d99e919        172.19.3.254:8787/rhosp14/openstack-nova-api:14.0-109.1560457048                    "kolla_start"            About an hour ago   Up About an hour (unhealthy)                        nova_metadata
bd00770fca9d        172.19.3.254:8787/rhosp14/openstack-nova-api:14.0-109.1560457048                    "kolla_start"            About an hour ago   Up About an hour (healthy)                          nova_api
~~~

In OSP 13, this looks like this:
~~~
[root@overcloud-controller-0 ~]# docker exec -it nova_metadata cat /openstack/healthcheck
#!/bin/sh

. ${HEALTHCHECK_SCRIPTS:-/usr/share/openstack-tripleo-common/healthcheck}/common.sh

check_url=$(get_url_from_vhost /etc/httpd/conf.d/10-nova_api_wsgi.conf)
healthcheck_curl ${check_url}
~~~

Within the OSP 14 container, I see:
~~~
()[nova@overcloud-controller-0 /]$ cat /openstack/healthcheck 
#!/bin/sh

. ${HEALTHCHECK_SCRIPTS:-/usr/share/openstack-tripleo-common/healthcheck}/common.sh


if pgrep -f nova-metadata; then
    check_url=$(get_url_from_vhost /etc/httpd/conf.d/10-nova_metadata_wsgi.conf)
else
    check_url=$(get_url_from_vhost /etc/httpd/conf.d/10-nova_api_wsgi.conf)
fi

healthcheck_curl ${check_url}
~~~

However, the test in the if condition is bad, because:
~~~
()[nova@overcloud-controller-0 /]$ pgrep -f nova-metadata
()[nova@overcloud-controller-0 /]$ ps aux | grep nova-metadat
nova         287  0.0  0.0   9088   668 ?        S+   16:49   0:00 grep nova-metadat
()[nova@overcloud-controller-0 /]$ ps aux | grep nova
nova           1  1.6  1.1 341896 94836 ?        Ss   16:42   0:06 /usr/bin/python2 /usr/bin/nova-api-metadata
nova         240  0.1  0.0  11816  1780 ?        Ss   16:48   0:00 /bin/bash
nova         294  0.0  0.0  51740  1740 ?        R+   16:49   0:00 ps aux
nova         295  0.0  0.0   9088   672 ?        S+   16:49   0:00 grep nova
()[nova@overcloud-controller-0 /]$ 
~~~

However, within the nova-metadata container on my system, I see the API file as well, and hence the health check reports as healthy:
~~~
()[nova@overcloud-controller-0 /]$ ls /etc/httpd/conf.d/10-nova_api_wsgi.conf
/etc/httpd/conf.d/10-nova_api_wsgi.conf
~~~

But in a customer environment, this fails on the healthcheck ...