openstack/fuel-docs
Code Issues Proposed changes

Describes Ceilometer support for vCenter in 6.1

Browse Source 
In 6.1 release, Ceilometer agent is installed on the same node
as Controller and Compute.
This provides metrics collection for the instances.

The issue will be covered in the corresponding
section of User Guide.

Change-Id: I083e149a598b8fb6a5422ffb5f2c3f2fd9b52072
Implements: blueprint ceilometer-support-for-vcenter
This commit is contained in:
Irina Povolotskaya
2015-04-29 16:32:12 +03:00
parent c1b954f6a8
commit 68fd75304f
2 changed files with 28 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
pages/operations/7800-vcenter.rst
View File

@@ -23,6 +23,8 @@ and/or the Nova CLI:
- Use the Nova CLI to see the VMware cluster resources
or to boot a new VM in vCenter.
.. _1-1 mapping:
Nova-compute and vSphere clusters mapping
-----------------------------------------

26
pages/user-guide/7300-vcenter.rst
View File

@@ -111,6 +111,32 @@ metrics are collected for the vCenter environment.
For more details about the Ceilometer plugin for vCenter,
see `Support for VMware vCenter Server <https://wiki.openstack.org/wiki/Ceilometer/blueprints/vmware-vcenter-server#Support_for_VMware_vCenter_Server>`_.
Previously, when you selected vCenter,
Compute and Controller roles were assigned to the same node,
while Ceilometer compute agent was not present on this node.
In Fuel 6.1 release, this logic was changed to provide
metrics collection for the instances.
For previous Fuel releases,
support for collecting polling meters
from instances on vCenter was not implemented:
only central agent polled services like Glance and Swift on
Controller node. No metrics was collected from compute nodes.
In 6.1, Ceilometer support for vCenter is implemented according
to 1-1 mapping principle (the one done between :ref:`nova-compute and
vSphere cluster <1-1 mapping>`).
Now Ceilometer compute service is available
for each vSphere cluster. That means,every agent polls resources
about instances from those that only relate to their vSphere cluster.
Every agent uses its own configuration file with authentication
parameters for its specific vSphere cluster.
What is more, monitoring under Pacemaker is introduced
for every Ceilometer compute service to avoid failures
(for example, failover of primary controller node)
and missing polling data as the result.
.. raw: pdf
PageBreak