OVN DB Tables Rows Count Metrics

This patch introduces the following changes. 1. Collectd exec plugins have been created to capture OVN DB tables rows count numbers from the controller host. 2. A Grafana panel has been created for OVN DB Tables Rows Count metrics. Change-Id: I274cc71d87ef79aefd381e690a13de6cb2c9b45e
2021-09-23 14:19:19 +05:30 · 2021-09-23 14:19:19 +05:30 · 42a7cf1b6d
commit 42a7cf1b6d
parent aaa9db0ec6
9 changed files with 199 additions and 0 deletions
--- a/ansible/install/group_vars/all.yml
+++ b/ansible/install/group_vars/all.yml
@ -311,6 +311,14 @@ ovsagent_controller_monitor: false
 ovsagent_networker_monitor: false
 ovs_flows_monitoring: false

+######################################
+# OVN DB Tables Rows Count Monitoring
+######################################
+# Monitors OVN DB Tables rows count on controller.
+# You have to add wheel to sudo list with no password prompt on the controller host,
+# before enabling this plugin.
+ovn_monitoring: false
+
 controller_monitored_ints:
  - "tap"

--- a/ansible/install/roles/collectd-openstack/tasks/main.yml
+++ b/ansible/install/roles/collectd-openstack/tasks/main.yml
@ -256,6 +256,8 @@
  with_items:
    - src: ovs_flows.sh
      dest: /usr/local/bin/ovs_flows.sh
+    - src: ovn_monitoring.sh
+      dest: /usr/local/bin/ovn_monitoring.sh
  when: "('Controller' in group_names ) or ('Compute' in group_names)"

 # Rabbitmq monitoring
--- a/ansible/install/roles/collectd-openstack/templates/controller.collectd.conf.j2
+++ b/ansible/install/roles/collectd-openstack/templates/controller.collectd.conf.j2
@ -133,6 +133,73 @@ PreCacheChain "PreCache"
 </Plugin>
 {% endif %}

+{% if ovn_monitoring %}
+<Plugin exec>
+  Instance "ovn_sbdb_Address_Set"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Address_Set"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Port_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Port_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Logical_Flow"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Logical_Flow"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Multicast_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Multicast_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Datapath_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Datapath_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Port_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Port_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_MAC_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "MAC_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Switch"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Switch"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Switch_Port"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Switch_Port"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Address_Set"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Address_Set"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Port_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Port_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Load_Balancer"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Load_Balancer"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_ACL"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "ACL"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router_Port"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router_Port"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router_Static_Route"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router_Static_Route"
+</Plugin>
+{% endif %}
+
 <Plugin ceph>
  LongRunAvgLatency false
  ConvertSpecialMetricTypes true
--- a/ansible/install/roles/collectd/tasks/main.yml
+++ b/ansible/install/roles/collectd/tasks/main.yml
@ -83,6 +83,13 @@
    dest: "{{ browbeat_containers_path }}/collectd-openstack/config/collectd.conf"
  when: "config_type == 'compute' and ovn_compute_collectd_plugin"

+- name: Add stack user to wheel
+  shell: |
+    podman exec -it -u root collectd-controller usermod -G wheel stack
+  become: yes
+  become_user: root
+  when: "config_type == 'controller' and ovn_monitoring"
+
 - name: Build and Run container
  block:
    - name: Build collectd-openstack container
@ -104,6 +111,10 @@
          {% if ovs_flows_monitoring %}
          -v /etc/openvswitch/:/etc/openvswitch/ -v /var/run/openvswitch/:/var/run/openvswitch/ \
          {% endif %}
+          {% if ovn_monitoring %}
+          -v /var/lib/openvswitch/ovn/ovnnb_db.sock:/var/lib/openvswitch/ovn/ovnnb_db.sock \
+          -v /var/lib/openvswitch/ovn/ovnsb_db.sock:/var/lib/openvswitch/ovn/ovnsb_db.sock \
+          {% endif %}
          {% endif %}
          collectd-openstack
      become: yes
--- a/ansible/install/roles/collectd/templates/controller.collectd.conf.j2
+++ b/ansible/install/roles/collectd/templates/controller.collectd.conf.j2
@ -454,6 +454,73 @@ PreCacheChain "PreCache"
 </Plugin>
 {% endif %}

+{% if ovn_monitoring %}
+<Plugin exec>
+  Instance "ovn_sbdb_Address_Set"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Address_Set"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Port_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Port_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Logical_Flow"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Logical_Flow"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Multicast_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Multicast_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Datapath_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Datapath_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_Port_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "Port_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_sbdb_MAC_Binding"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "sb" "MAC_Binding"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Switch"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Switch"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Switch_Port"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Switch_Port"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Address_Set"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Address_Set"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Port_Group"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Port_Group"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Load_Balancer"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Load_Balancer"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_ACL"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "ACL"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router_Port"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router_Port"
+</Plugin>
+<Plugin exec>
+  Instance "ovn_nbdb_Logical_Router_Static_Route"
+  Exec stack "/usr/local/bin/ovn_monitoring.sh" "nb" "Logical_Router_Static_Route"
+</Plugin>
+{% endif %}
+
 {%if iostat_controller_collectd_plugin %}
 <Plugin python>
  ModulePath "/usr/local/bin/"
--- a/ansible/install/roles/grafana-dashboards/templates/openstack_general_system_performance.yaml.j2
+++ b/ansible/install/roles/grafana-dashboards/templates/openstack_general_system_performance.yaml.j2
@ -164,6 +164,8 @@ dashboard:
  {% include 'partials/osp_resources.yaml' %}

  {% include 'partials/osp_response_times.yaml' %}
+
+  {% include 'partials/ovn_db_tables.yaml' %}
 {% endif %}

 {% include 'partials/ovn_metrics.yaml' %}
--- a/ansible/install/roles/grafana-dashboards/templates/partials/ovn_db_tables.yaml
+++ b/ansible/install/roles/grafana-dashboards/templates/partials/ovn_db_tables.yaml
@ -0,0 +1,33 @@
+    - title: OVN DB Tables Rows Count
+      collapse: true
+      height: 200px
+      showTitle: true
+      panels:
+        - title: $Cloud - $Node - OVN DB Tables Rows Count
+          type: graph
+          legend:
+            alignAsTable: true
+            avg: false
+            current: true
+            max: true
+            min: true
+            rightSide: true
+            show: true
+            total: false
+            values: true
+          nullPointMode: 'null'
+          targets:
+            - target: alias($Cloud.$Node.ovn-ACL.gauge-ovn_ACL, 'ovn_ACL')
+            - target: alias($Cloud.$Node.ovn-Address_Set.gauge-ovn_Address_Set, 'ovn_Address_Set')
+            - target: alias($Cloud.$Node.ovn-Datapath_Binding.gauge-ovn_Datapath_Binding, 'ovn_Datapath_Binding')
+            - target: alias($Cloud.$Node.ovn-Load_Balancer.gauge-ovn_Load_Balancer, 'ovn_Load_Balancer')
+            - target: alias($Cloud.$Node.ovn-Logical_Flow.gauge-ovn_Logical_Flow, 'ovn_Logical_Flow')
+            - target: alias($Cloud.$Node.ovn-Logical_Router.gauge-ovn_Logical_Router, 'ovn_Logical_Router')
+            - target: alias($Cloud.$Node.ovn-Logical_Router_Port.gauge-ovn_Logical_Router_Port, 'ovn_Logical_Router_Port')
+            - target: alias($Cloud.$Node.ovn-Logical_Router_Static_Route.gauge-ovn_Logical_Router_Static_Route, 'ovn_Logical_Router_Static_Route')
+            - target: alias($Cloud.$Node.ovn-Logical_Switch.gauge-ovn_Logical_Switch, 'ovn_Logical_Switch')
+            - target: alias($Cloud.$Node.ovn-Logical_Switch_Port.gauge-ovn_Logical_Switch_Port, 'ovn_Logical_Switch_Port')
+            - target: alias($Cloud.$Node.ovn-MAC_Binding.gauge-ovn_MAC_Binding, 'ovn_MAC_Binding')
+            - target: alias($Cloud.$Node.ovn-Multicast_Group.gauge-ovn_Multicast_Group, 'ovn_Multicast_Group')
+            - target: alias($Cloud.$Node.ovn-Port_Binding.gauge-ovn_Port_Binding, 'ovn_Port_Binding')
+            - target: alias($Cloud.$Node.ovn-Port_Group.gauge-ovn_Port_Group, 'ovn_Port_Group')
--- a/browbeat-containers/collectd-openstack/Dockerfile
+++ b/browbeat-containers/collectd-openstack/Dockerfile
@ -27,6 +27,7 @@ ADD files/collectd_gnocchi_status.py /usr/local/bin/collectd_gnocchi_status.py
 ADD files/collectd_rabbitmq_monitoring.py /usr/local/bin/collectd_rabbitmq_monitoring.py
 ADD files/collectd_swift_stat.py /usr/local/bin/collectd_swift_stat.py
 ADD files/ovs_flows.sh /usr/local/bin/ovs_flows.sh
+ADD files/ovn_monitoring.sh /usr/local/bin/ovn_monitoring.sh

 ADD config/collectd.conf /etc/collectd.conf

--- a/browbeat-containers/collectd-openstack/files/ovn_monitoring.sh
+++ b/browbeat-containers/collectd-openstack/files/ovn_monitoring.sh
@ -0,0 +1,8 @@
+#!/bin/sh
+HOSTNAME="${COLLECTD_HOSTNAME:-`hostname -f`}"
+INTERVAL="${COLLECTD_INTERVAL:-15}"
+
+while sleep "$INTERVAL"; do
+  VALUE=$(sudo ovsdb-client dump unix:/var/lib/openvswitch/ovn/ovn$1_db.sock $2 | wc -l)
+  echo "PUTVAL \"$HOSTNAME/ovn-$2/gauge-ovn_$2\" interval=$INTERVAL N:$VALUE"
+done