From c6a9db0f744cd38d05a7da4adb692472732ab1ab Mon Sep 17 00:00:00 2001
From: Meg McRoberts <mmcroberts@mirantis.com>
Date: Sun, 28 Dec 2014 18:24:21 -0800
Subject: [PATCH] RN-6.0 -- Ceilometer last minute bugs for Release Notes

Added two last-minute Ceilometer bugs to the Release Notes.

Change-Id: I7e9121c071feca82099f3a71112d06fa6e20455a
---
 .../v6-0/other/5010-ceilometer-mongodb.rst    | 59 ++++++++++++++-----
 1 file changed, 44 insertions(+), 15 deletions(-)

diff --git a/pages/release-notes/v6-0/other/5010-ceilometer-mongodb.rst b/pages/release-notes/v6-0/other/5010-ceilometer-mongodb.rst
index b9ed417d7..d845e1a20 100644
--- a/pages/release-notes/v6-0/other/5010-ceilometer-mongodb.rst
+++ b/pages/release-notes/v6-0/other/5010-ceilometer-mongodb.rst
@@ -7,15 +7,17 @@ OpenStack Telemetry (Ceilometer) and MongoDB Database
 New Features and Resolved Issues in Mirantis OpenStack 6.0
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 
-* Ceilometer now successfully connects to AMQP after primary controller is shut down.
+* Ceilometer now successfully connects
+  to AMQP after the primary controller is shut down.
   See `LP1373569 <https://bugs.launchpad.net/fuel/+bug/1373569>`_.
 
-* Logs from MongoDB server are now written to a separate file instead of syslog
-  to avoid free disk space problems.
+* Logs from the MongoDB server are now written
+  to a separate file instead of syslog
+  to prevent free disk space problems.
   See `LP1367234 <https://bugs.launchpad.net/fuel/+bug/1367234>`_.
 
-* MongoDB provisioning no longer fails; it sets up a cluster
-  in recovery state.
+* MongoDB provisioning no longer fails;
+  it sets up a cluster in recovery state.
   See `LP1381826 <https://bugs.launchpad.net/fuel/+bug/1381826>`_.
 
 * MongoDB now retries on AutoReconnect exceptions.
@@ -47,14 +49,41 @@ MongoDB issues
   Ceilometer randomly fails in HA mode.
   See `LP1371799 <https://bugs.launchpad.net/fuel/+bug/1371799>`_.
 
-Ceilometer does not collect some notifications for Swift
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Ceilometer issues
+~~~~~~~~~~~~~~~~~
+
+* When you deploy an HA-mode environment,
+  create a container and object and download the object,
+  Ceilometer does not collect some
+  `metrics <http://docs.openstack.org/developer/ceilometer/measurements.html>`_
+  (such as *storage.objects.incoming.bytes*,
+  *storage.objects.outgoing.bytes*, and *storage.api.request*)
+  for :ref:`Swift<swift-object-storage-term>`.
+  See `LP1400240 <https://bugs.launchpad.net/bugs/1400240>`_.
+
+* The `ceilometer-agent-notification` service
+  may fail to connect to the queue after RabbitMQ is restarted.
+  This seems to happen only in HA mode
+  and we have only been able to reproduce one instance
+  of `ceilometer-agent-notification` failing.
+  As a work around, restart `ceilometer-agent-notification`
+  after restarting RabbitMQ.
+  See `LP1393505 <https://bugs.launchpad.net/mos/+bug/1393505>`_.
+
+* Ceilometer does not obtain Nova \*.rate pollsters.
+  This is related to an upstream bug:
+  `1394228 <https://bugs.launchpad.net/ceilometer/+bug/1394228>`_.
+  See `LP1400324 <https://bugs.launchpad.net/mos/+bug/1400324>`_.
+
+  As a work-around:
+
+  #. Add `Patch 139037 <https://review.openstack.org/#/c/139037/>`_
+     to the Ceilometer code on all nodes in the environment.
+
+  #. Add `Patch 147180 <http://paste.openstack.org/show/147180/>`_
+     to `sinks` in the */etc/ceilometer/pipeline.yaml* file
+     on all nodes in the environment.
+
+  #. Restart Ceilometer on all nodes in the environment.
+
 
-When you deploy HA-mode environment,
-create a container and object and download the object,
-Ceilometer does not obtain some
-`metrics <http://docs.openstack.org/developer/ceilometer/measurements.html>`_
-(such as *storage.objects.incoming.bytes*,
-*storage.objects.outgoing.bytes*, and *storage.api.request*)
-for :ref:`Swift<swift-object-storage-term>`.
-See `LP1400240 <https://bugs.launchpad.net/bugs/1400240>`_.