Ceph collectd plugin

+ Monitor the Ceph Mon socket on Controller-0
+ Monitor 1st Ceph OSD socket on each CephStorage Node
+ Graphs for Ceph Mon are included on the General Dashboard
+ Ceph Dashboard covers Ceph Processes + Ceph Mon Metrics
+ Small adjustments and improvements to dashboards
  + Stepped lines for queues
  + Try to use 200px for information denisty

* OSD metrics are collected but not graphed yet *

Change-Id: I97d65b98b5204e924eb5167adfba53c2c754214c

ansible/install/grafana-dashboards.yml

@@ -14,6 +14,7 @@
     cloud_dashboards: true
     overwrite_existing: true
     static_dashboards:
+      - cloud_ceph_monitoring
       - cloud_gnocchi_status
       - cloud_instance_count
       - cloud_keystone_token_count

ansible/install/group_vars/all.yml

@@ -91,6 +91,13 @@ apache_undercloud_mod_status_port: 5001
 apache_controller_collectd_plugin: false
 apache_controller_mod_status_port: 5001
 
+########################
+# Ceph plugin
+########################
+# Overcloud Controller
+ceph_controller_collectd_plugin: false
+ceph_storage_collectd_plugin: false
+
 ########################
 # Gnocchi backlog plugin
 ########################

ansible/install/roles/collectd-openstack/tasks/main.yml

@@ -25,6 +25,7 @@
   with_items:
     - collectd
     - collectd-apache
+    - collectd-ceph
     - collectd-mysql
     - collectd-turbostat
 
@@ -156,6 +157,15 @@
   when:  "('undercloud' in group_names or '{{inventory_hostname}}' == groups['controller'][0])"
 # End Rabbitmq monitoring
 
+# CephStorage OSD monitoring
+# Only Monitors a single OSD on each CephStorage Node
+- name: Get 1st OSD socket
+  shell: ls /var/run/ceph/ceph-osd.*.asok | head -n 1 | egrep -o '[0-9]+'
+  register: cephstorage_osd_socket
+  become: true
+  when: "('cephstorage' in group_names and {{ceph_storage_collectd_plugin}} == true)"
+# End CephStorage OSD monitoring
+
 - name: Configure collectd.conf
   template:
     src: "{{config_type}}.collectd.conf.j2"

ansible/install/roles/collectd-openstack/templates/cephstorage.collectd.conf.j2

@@ -16,6 +16,9 @@ LoadPlugin "logfile"
 </Plugin>
 
 LoadPlugin write_graphite
+{% if ceph_storage_collectd_plugin %}
+LoadPlugin ceph
+{% endif %}
 LoadPlugin cpu
 LoadPlugin conntrack
 LoadPlugin df
@@ -75,6 +78,16 @@ PreCacheChain "PreCache"
   </Carbon>
 </Plugin>
 
+{% if ceph_storage_collectd_plugin %}
+<Plugin ceph>
+  LongRunAvgLatency false
+  ConvertSpecialMetricTypes true
+  <Daemon "osd.{{inventory_hostname}}.{{cephstorage_osd_socket.stdout}}">
+    SocketPath "/var/run/ceph/ceph-osd.{{cephstorage_osd_socket.stdout}}.asok"
+  </Daemon>
+</Plugin>
+
+{% endif %}
 <Plugin df>
   ValuesPercentage true
 </Plugin>

ansible/install/roles/collectd-openstack/templates/controller.collectd.conf.j2

@@ -19,6 +19,11 @@ LoadPlugin write_graphite
 {% if apache_controller_collectd_plugin %}
 LoadPlugin apache
 {% endif %}
+{% if ceph_controller_collectd_plugin %}
+{% if inventory_hostname == groups['controller'][0] %}
+LoadPlugin ceph
+{% endif %}
+{% endif %}
 LoadPlugin cpu
 LoadPlugin conntrack
 {% if keystone_overcloud_collectd_plugin %}
@@ -96,6 +101,19 @@ PreCacheChain "PreCache"
   </Instance>
 </Plugin>
 {% endif %}
+{% if ceph_controller_collectd_plugin %}
+{% if inventory_hostname == groups['controller'][0] %}
+
+<Plugin ceph>
+  LongRunAvgLatency false
+  ConvertSpecialMetricTypes true
+  <Daemon "mon.{{inventory_hostname}}">
+    SocketPath "/var/run/ceph/ceph-mon.{{inventory_hostname}}.asok"
+  </Daemon>
+</Plugin>
+
+{% endif %}
+{% endif %}
 {% if keystone_overcloud_collectd_plugin %}
 {%if inventory_hostname == groups['controller'][0] %}
 

ansible/install/roles/grafana-dashboards/files/cloud_gnocchi_status.json

@@ -51,7 +51,7 @@
             "seriesOverrides": [],
             "span": 12,
             "stack": false,
-            "steppedLine": false,
+            "steppedLine": true,
             "targets": [
               {
                 "refId": "A",
@@ -151,7 +151,7 @@
       ]
     },
     "time": {
-      "from": "now-24h",
+      "from": "now-1h",
       "to": "now"
     },
     "timepicker": {

ansible/install/roles/grafana-dashboards/files/cloud_keystone_token_count.json

@@ -151,7 +151,7 @@
       ]
     },
     "time": {
-      "from": "now-24h",
+      "from": "now-1h",
       "to": "now"
     },
     "timepicker": {

ansible/install/roles/grafana-dashboards/files/cloud_rabbitmq_monitoring.json

@@ -11,7 +11,7 @@
     "links": [],
     "rows": [
       {
-        "collapse": false,
+        "collapse": true,
         "height": "200px",
         "panels": [
           {
@@ -479,8 +479,8 @@
         "titleSize": "h6"
       },
       {
-        "collapse": false,
-        "height": 250,
+        "collapse": true,
+        "height": "200px",
         "panels": [
           {
             "aliasColors": {},
@@ -596,7 +596,7 @@
             "seriesOverrides": [],
             "span": 12,
             "stack": false,
-            "steppedLine": false,
+            "steppedLine": true,
             "targets": [
               {
                 "refId": "A",
@@ -880,7 +880,7 @@
             "seriesOverrides": [],
             "span": 12,
             "stack": false,
-            "steppedLine": false,
+            "steppedLine": true,
             "targets": [
               {
                 "refId": "A",
@@ -932,7 +932,7 @@
         "titleSize": "h6"
       },
       {
-        "collapse": false,
+        "collapse": true,
         "height": "200px",
         "panels": [
           {

ansible/install/roles/grafana-dashboards/vars/main.yml

@@ -339,8 +339,8 @@ per_process_panels:
   OpenStack-CephStorage:
     - name: "Ceph"
       processes:
-        - ceph-osd
         - ceph-mon
+        - ceph-osd
     - name: "Open vSwitch"
       processes:
         - ovs-vswitchd
@@ -399,8 +399,8 @@ per_process_panels:
         - ceilometer_wsgi
     - name: "Ceph"
       processes:
-        - ceph-osd
         - ceph-mon
+        - ceph-osd
     - name: "Cinder"
       processes:
         - cinder-api