x
/
fuel-plugin-lma-collector
Code Issues Proposed changes

Remove obsolete documentation 

This change removes the references to the service status metrics that
existed in the 0.7 release.

Change-Id: Iac11c6ae4c242ed5330734bcf517c7ab45617579
(cherry picked from commit 40a3c1b6d1)
This commit is contained in:
Simon Pasquier 2015-11-13 15:27:11 +01:00
parent 36cd12d10c
commit 929a5c186f
1 changed files with 0 additions and 94 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

94
doc/source/dev/metrics/openstack.rst
View File

@ -59,14 +59,6 @@ These metrics are retrieved from the Nova database.
'conductor', 'scheduler', 'cert' or 'consoleauth') and a ``state`` field (one
of 'up', 'down' or 'disabled').
Status metrics, see :ref:`service status <service_status>` for details:
* ``openstack_nova_services_status``, status of Nova services (workers)
computed from ``openstack_nova_services``.
* ``openstack_nova_pool_status``, status of API services located behind the HAProxy load-balancer,
computed from ``haproxy_backend_servers``.
* ``openstack_nova_status``, the global status of the Nova service.
Identity
^^^^^^^^
@ -78,12 +70,6 @@ These metrics are retrieved from the Keystone API.
* ``openstack_keystone_users``, the number of users by state. The metric
contains a ``state`` field (either 'enabled' or 'disabled').
Status metrics, see :ref:`service status <service_status>` for details:
* ``openstack_keystone_pool_status``, status of API services located behind the
HAProxy load-balancer, computed from ``haproxy_backend_servers``.
* ``openstack_keystone_status``, the global status of the Keystone service.
Volume
^^^^^^
@ -110,13 +96,6 @@ These metrics are retrieved from the Cinder database.
The metric contains a ``service`` field (one of 'volume', 'backup',
'scheduler') and a ``state`` field (one of 'up', 'down' or 'disabled').
Status metrics, see :ref:`service status <service_status>` for details:
* ``openstack_cinder_services_status``, status of Cinder services (workers) computed from ``openstack_cinder_services``.
* ``openstack_cinder_pool_status``, status of API services located behind the HAProxy load-balancer,
computed from ``haproxy_backend_servers``.
* ``openstack_cinder_status``, the global status of the Cinder.
Image
^^^^^
@ -135,12 +114,6 @@ These metrics are retrieved from the Glance API.
``state`` is one of 'queued', 'saving', 'active', 'killed', 'deleted',
'pending_delete'. ``visibility`` is either 'public' or 'private'.
Status metrics, see :ref:`service status <service_status>` for details:
* ``openstack_glance_pool_status``, status of the API service located behind the HAProxy load-balancer,
computed from ``haproxy_backend_servers``.
* ``openstack_glance_status``, the global status of the Glance service.
Network
^^^^^^^
@ -164,76 +137,9 @@ These metrics are retrieved from the Neutron database.
and state. The metric contains ``service`` (one of 'dhcp', 'l3', 'metadata'
or 'openvswitch') and ``state`` (one of 'up', 'down' or 'disabled') fields.
Status metrics, see :ref:`service status <service_status>` for details:
* ``openstack_neutron_agents_status``, status of Neutron services (workers) computed from metric ``openstack_neutron_agents``.
* ``openstack_neutron_pool_neutron_status``, status of the API service located behind the HAProxy load-balancer,
computed from ``haproxy_backend_servers``.
* ``openstack_neutron_status``, the global status of the Neutron service.
API response times
^^^^^^^^^^^^^^^^^^
* ``openstack_<service>_http_responses``, the time (in second) it took to serve the HTTP request. The metric contains ``http_method`` (eg 'GET', 'POST', and so on) and ``http_status`` (eg '200', '404', and so on) fields.
``<service>`` is one of 'cinder', 'glance', 'heat' 'keystone', 'neutron' or 'nova'.
Service status
^^^^^^^^^^^^^^
.. _service_status:
.. note:: This section is obsolete and should be rewritten.
A **global status** is computed for each OpenStack service (``openstack.<service>.status``),
where the value is one of:
* 0, meaning OKAY
* 1, meaning WARN
* 2, meaning FAIL
* 3, meaning UNKNOWN (no metric to determine the status)
The **global status** of a service is based on its **underlying status**,
where the value is one of:
* 0, meaning UP
* 1, meaning DEGRADED
* 2, meaning DOWN
* 3, meaning UNKNOWN (no metric to determine the status)
**Underlying status**:
* ``openstack.<service>.endpoint.<api>.status``, status of all API of the service,
based on related :ref:`service checks <service_checks>` (``openstack.<api>.check_api``).
To notice that the endpoint status cannot be DEGRADED.
* ``openstack.<service>.pool.<backend>.status``, status of all HAproxy backend pools,
based on related status of :ref:`HAproxy server states <haproxy_backend_metric>` (``haproxy.backend.<backend>.servers.(up|down)``).
The status is
* OKAY if all servers are UP.
* DEGRADED if one or more servers are DOWN and at least one server is UP.
* DOWN if all servers are DOWN.
Furhtermore, the global statutes of the *compute*, *volume* and *network* services
are also based respectively on these underlying 'worker' status:
* ``openstack.nova.services.<service>.status``, status of Nova services computed from ``openstack.nova.services.<service>.<service_state>``,
see :ref:`Nova service states <compute-service-state-metrics>`.
* ``openstack.cinder.services.<service>.status``, status of Nova services computed from ``openstack.cinder.services.<service>.<service_state>``,
see :ref:`Cinder service states <volume-service-state-metrics>`.
* ``openstack_neutron.agents.<agent_type>.status``, status of Neutron agents computed from ``openstack.neutron.agents.<agent_type>.<agent_state>``,
see :ref:`Neutron agent states <network-agent-state-metrics>`.
The status of these 3 above is determined as follow:
* OK if all workers are UP and there is no worker DOWN, note that DISABLED workers are ignored.
* DEGRADED if one or more workers are DOWN and at least one worker is UP.
* DOWN if there is no UP worker.
The **global status** determination follows these simple rules:
* OK if all underlying status are OK.
* WARN if one of underlying status is DEGRADED.
* FAIL if one of underlying status is DOWN.
* UNKNOWN if one of underlying status is UNKNOWN.