Prometheus linting warnings

[JamesGibo]

Running the output of openstack exporter through the Prometheus metrics checking tool returns a number of warnings that may want to be changed in future versions.

curl https://<hostname>:9180/probe?cloud=default | promtool check metrics

openstack_cinder_agent_state counter metrics should have "_total" suffix
openstack_cinder_agent_state label names should be written in 'snake_case' not 'camelCase'
openstack_cinder_limits_volume_max_gb metric names should not contain abbreviated units
openstack_cinder_limits_volume_used_gb metric names should not contain abbreviated units
openstack_cinder_pool_capacity_free_gb metric names should not contain abbreviated units
openstack_cinder_pool_capacity_total_gb metric names should not contain abbreviated units
openstack_cinder_volume_status_counter metric name should not include type 'counter'
openstack_neutron_agent_state counter metrics should have "_total" suffix
openstack_neutron_agent_state label names should be written in 'snake_case' not 'camelCase'
openstack_neutron_network_ip_availabilities_total non-counter metrics should not have "_total" suffix
openstack_nova_agent_state counter metrics should have "_total" suffix
openstack_nova_agent_state label names should be written in 'snake_case' not 'camelCase'

[niedbalski]

Hey @JamesGibo sorry for taking so much time to attend your report. Thank you so much. Question, are you interested to work on 1) adding a ci tool to check this? 2) fixing some of those? I am more than happy to guide you through the review/process.

[cityofships]

I'm happy to contribute to the project to resolve the issue, just:
1) Should be implemented after or as a part of #178 ?
2) This will be a breaking change.