[stein] Prometheus OpenStack Exporter Broken

[Joseph M]

**Bug Report**

What happened:
Two issues:
1. The clouds.yml template provided to prometheus-openstack-exporter is rendered with the wrong auth_url for multi region environments, using kolla_internal_fqdn instead of keystone_internal_url as described by the Kolla-Ansible multi region guide.

https://opendev.org/openstack/kolla-ansible/src/branch/master/ansible/roles/prometheus/templates/clouds.yml.j2

2. The openstack-exporter version shipped in kolla stable/stein is 0.0.7 which does not retrieve stats for a standard stein deployment:

time="2019-09-21T01:26:21-04:00" level=info msg="Fetching list of services" source="nova.go:53"
time="2019-09-21T01:26:21-04:00" level=info msg="Fetching domains information" source="keystone.go:45"
time="2019-09-21T01:26:21-04:00" level=info msg="Fetching volumes info" source="cinder.go:54"
time="2019-09-21T01:26:21-04:00" level=info msg="Fetching images list" source="glance.go:39"
time="2019-09-21T01:26:21-04:00" level=info msg="Fetching floating ips list" source="neutron.go:47"
time="2019-09-21T01:26:21-04:00" level=error msg="Get /volumes: Get /volumes: unsupported protocol sch eme \"\"" source="cinder.go:57"
time="2019-09-21T01:26:21-04:00" level=info msg="Fetching services state information" source="cinder.g o:60"
time="2019-09-21T01:26:21-04:00" level=error msg="Get /os-services: Get /os-services: unsupported prot ocol scheme \"\"" source="cinder.go:63"
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x8 pc=0x82827d]

goroutine 81 [running]:
main.(*CinderExporter).Collect(0xc00001cac0, 0xc0002863c0)
        /app/cinder.go:66 +0x1dd
github.com/prometheus/client_golang/prometheus.(*Registry).Gather.func1()
        /app/vendor/github.com/prometheus/client_golang/prometheus/registry.go:434 +0x193
created by github.com/prometheus/client_golang/prometheus.(*Registry).Gather
        /app/vendor/github.com/prometheus/client_golang/prometheus/registry.go:526 +0xe23

Setting "prometheus_openstack_exporter_tag: master" fixes the above issue due to the newer exporter version.

What you expected to happen:
It renders the clouds.yml template with the correct auth_url value and pulls stats correctly.

How to reproduce it (minimal and precise):
Set up a multi-region environment with the multi-region guide and set "enable_prometheus: yes"
https://docs.openstack.org/kolla-ansible/latest/user/multi-regions.html

**Environment**:
* OS (e.g. from /etc/os-release): CentOS Linux release 7.6.1810 (Core)
* Kernel (e.g. `uname -a`): 3.10.0-957.21.3.el7.x86_64
* Docker version if applicable (e.g. `docker version`): 19.03.2
* Kolla-Ansible version (e.g. `git head or tag or stable branch` or pip package version if using release): 8.0.2.dev4
* Docker image Install type (source/binary): source
* Docker image distribution: centos
* Are you using official images from Docker Hub or self built? Official
* If self built - Kolla version and environment used to build: NA
* Share your inventory file, globals.yml and other configuration files if relevant

[Mark Goddard (mgoddard)]

1. Makes sense, please propose a fix.

2. I have seen this raised a few times. There has been a suggestion of bumping the openstack exporter to 0.0.9 on stable/stein. This fixes a common issue, and I think is in line with our stable branch policy since it's not a major version bump.

[Mark Goddard (mgoddard)]

Here is a bug for the cinder issue: https://bugs.launchpad.net/kolla-ansible/+bug/1838391

[Marcin Juszkiewicz (hrw)]

https://review.opendev.org/#/c/689437/ bumped OpenStack Exporter to 0.6.0 version (in train). https://review.opendev.org/#/c/689659/ does same for stein.