openstack/nova
Code Issues Proposed changes

doc: add dev policy about no new metrics monitors

Browse Source 
At the newton midcycle we talked about the proposed
memory bandwidth monitor plugin and decided that we
really just don't want nova in the business of metrics
gathering, and are moving to deprecate that functionality.

There are several specs open related to adding more
metrics monitors and/or exposing existing ones, so this
codifies the policy on nova as metrics gatherer.

Change-Id: Ie47e91776bc36a22fd48d3673c69c842509489a1
This commit is contained in:
Matt Riedemann
2016-10-11 20:41:47 -04:00
parent 610293fdd5
commit 0a99fc1c4e
1 changed files with 29 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
doc/source/policies.rst
View File

@@ -101,3 +101,32 @@ procedure is:
design spec might be the best way to achieve this.
.. _`an enforced time-based solution`: https://lists.launchpad.net/openstack/msg08574.html
Metrics Gathering
=================
Nova currently has a monitor plugin to gather CPU metrics on compute nodes.
This feeds into the MetricsFilter and MetricsWeigher in the scheduler. The
CPU metrics monitor is only implemented for the libvirt compute driver.
External projects like `Ceilometer`_ and `Watcher`_ consume these metrics.
Over time people have tried to add new monitor plugins for things like memory
bandwidth. There have also been attempts to expose these monitors over CLI,
the REST API, and notifications.
At the `Newton midcycle`_ it was decided that Nova does a poor job as a metrics
gathering tool, especially as it's incomplete, not tested, and there are
numerous other tools available to get this information as their primary
function.
Therefore, there is a freeze on adding new metrics monitoring plugins which
also includes exposing existing monitored metrics outside of Nova, like with
the nova-manage CLI, the REST API, or the notification bus. Long-term, metrics
gathering will likely be deprecated within Nova. Since there is not yet a clear
replacement, the deprecation is open-ended, but serves as a signal that new
deployments should not rely on the metrics that Nova gathers and should instead
focus their efforts on alternative solutions for placement.
.. _Ceilometer: http://docs.openstack.org/developer/ceilometer/
.. _Watcher: http://docs.openstack.org/developer/watcher/
.. _Newton midcycle: http://lists.openstack.org/pipermail/openstack-dev/2016-August/100600.html