Merge pull request #18 from tcpcloud/stacklight

Stacklight

galera/files/collectd_mysql.conf

@@ -0,0 +1,14 @@
+<LoadPlugin mysql>
+  Globals false
+</LoadPlugin>
+
+<Plugin mysql>
+    <Database "mysql">
+        Host "localhost"
+        Socket "{{ plugin.socket }}"
+        User "{{ plugin.username }}"
+        Password "{{ plugin.password }}"
+        MasterStats false
+        SlaveStats false
+    </Database>
+</Plugin>

galera/files/collectd_mysql_check.conf

@@ -0,0 +1,9 @@
+Import "collectd_mysql_check"
+
+<Module "collectd_mysql_check">
+    Host "localhost"
+    Socket "{{ plugin.socket }}"
+    Username "{{ plugin.username }}"
+    Password "{{ plugin.password }}"
+</Module>
+

galera/files/collectd_mysql_status.conf

@@ -0,0 +1,44 @@
+<LoadPlugin dbi>
+  Globals false
+</LoadPlugin>
+
+<Plugin dbi>
+  <Query "wsrep_ready">
+    Statement "select replace(lower(VARIABLE_NAME), 'wsrep_', 'cluster.') as metric, replace(replace(VARIABLE_VALUE, 'ON', 1), 'OFF', 0) as value from information_schema.GLOBAL_STATUS where VARIABLE_NAME IN ('wsrep_ready', 'wsrep_connected');"
+    MinVersion 50000
+    <Result>
+      Type "gauge"
+      InstancesFrom "metric"
+      ValuesFrom "value"
+    </Result>
+  </Query>
+  <Query "wsrep_cluster_status">
+    Statement "select replace(replace(lower(VARIABLE_NAME), 'wsrep_', ''), '_', '.') as metric, replace(replace(replace(VARIABLE_VALUE, 'Primary', 1), 'Non-Primary', 2), 'Disconnected', 3) as value from information_schema.GLOBAL_STATUS where VARIABLE_NAME = 'wsrep_cluster_status';"
+    MinVersion 50000
+    <Result>
+      Type "gauge"
+      InstancesFrom "metric"
+      ValuesFrom "value"
+    </Result>
+  </Query>
+  <Query "wsrep_cluster">
+    Statement "select replace(replace(lower(VARIABLE_NAME), 'wsrep_', 'cluster.'), 'cluster_size', 'size') as metric, VARIABLE_VALUE as value from information_schema.GLOBAL_STATUS where VARIABLE_NAME IN ('wsrep_cluster_size', 'wsrep_replicated', 'wsrep_replicated_bytes', 'wsrep_received_bytes', 'wsrep_received', 'wsrep_local_commits', 'wsrep_local_cert_failures', 'wsrep_local_send_queue', 'Slow_queries');"
+    MinVersion 50000
+    <Result>
+      Type "gauge"
+      InstancesFrom "metric"
+      ValuesFrom "value"
+    </Result>
+  </Query>
+  <Database "mysql_status">
+    Driver "mysql"
+    DriverOption "mysql_unix_socket" "{{ plugin.socket }}"
+    DriverOption "username" "{{ plugin.username }}"
+    DriverOption "password" "{{ plugin.password }}"
+    DriverOption "dbname" "mysql"
+    SelectDB "mysql"
+    Query "wsrep_ready"
+    Query "wsrep_cluster_status"
+    Query "wsrep_cluster"
+  </Database>
+</Plugin>

galera/map.jinja

@@ -1,7 +1,7 @@
 
 {%- set master = salt['grains.filter_by']({
   'Debian': {
-    'pkgs': ['mysql-server-wsrep-5.6', 'galera-3', 'rsync', 'python-mysqldb', 'libmysqlclient18', 'psmisc', 'netcat', 'percona-xtrabackup', 'socat'],
+    'pkgs': ['mysql-server-wsrep-5.6', 'galera-3', 'rsync', 'python-mysqldb', 'libmysqlclient18', 'psmisc', 'netcat', 'percona-xtrabackup', 'socat', 'libdbd-mysql', 'python-pymysql'],
     'service': 'mysql',
     'wsrep_provider': '/usr/lib/galera/libgalera_smm.so',
     'log_file': '/var/log/mysql.log',
@@ -21,7 +21,7 @@
 
 {%- set slave = salt['grains.filter_by']({
   'Debian': {
-    'pkgs': ['mysql-server-wsrep', 'galera', 'rsync', 'python-mysqldb', 'libmysqlclient18', 'psmisc', 'netcat', 'percona-xtrabackup', 'socat'],
+    'pkgs': ['mysql-server-wsrep', 'galera', 'rsync', 'python-mysqldb', 'libmysqlclient18', 'psmisc', 'netcat', 'percona-xtrabackup', 'socat', 'libdbd-mysql', 'python-pymysql'],
     'service': 'mysql',
     'wsrep_provider': '/usr/lib/galera/libgalera_smm.so',
     'log_file': '/var/log/mysql.log',

galera/meta/collectd.yml

@@ -0,0 +1,29 @@
+{%- if pillar.galera is defined  %}
+{%- if pillar.galera.master is defined %}
+{%- from "galera/map.jinja" import master with context %}
+{%- set server = master %}
+{%- elif pillar.galera.slave is defined %}
+{%- from "galera/map.jinja" import slave with context %}
+{%- set server = slave %}
+{%- endif %}
+{%- endif %}
+
+{%- if server is defined %}
+local_plugin:
+  mysql:
+    template: galera/files/collectd_mysql.conf
+    socket: {{ server.socket }}
+    password: {{ server.admin.password }}
+    username: {{ server.admin.user }}
+  mysql_status:
+    template: galera/files/collectd_mysql_status.conf
+    socket: {{ server.socket }}
+    password: {{ server.admin.password }}
+    username: {{ server.admin.user }}
+  mysql_check:
+    plugin: python
+    template: galera/files/collectd_mysql_check.conf
+    socket: {{ server.socket }}
+    password: {{ server.admin.password }}
+    username: {{ server.admin.user }}
+{%- endif %}

galera/meta/grafana.yml

@@ -0,0 +1,4 @@
+dashboard:
+  galera:
+    format: json
+    template: galera/files/grafana_dashboards/mysql_influxdb.json

galera/meta/heka.yml

@@ -0,0 +1,80 @@
+log_collector:
+  decoder:
+    galera:
+      engine: sandbox
+      module_file: /usr/share/lma_collector/decoders/galera.lua
+      module_dir: /usr/share/lma_collector/common;/usr/share/heka/lua_modules
+      adjust_timezone: true
+  input:
+    galera:
+      engine: logstreamer
+      log_directory: "/var/log"
+      file_match: 'mysql/error.log$'
+      differentiator: ["galera"]
+      decoder: "galera_decoder"
+      splitter: "galera_splitter"
+  splitter:
+    galera:
+      engine: regex
+      delimiter: '\n([0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2})'
+      delimiter_eol: false
+metric_collector:
+  trigger:
+    mysql_check:
+      description: 'MySQL cannot be checked'
+      severity: down
+      rules:
+      - metric: mysql_check
+        relational_operator: '=='
+        threshold: 0
+        window: 60
+        periods: 0
+        function: last
+    mysql_node_connected:
+      description: 'The MySQL service has lost connectivity with the other nodes'
+      severity: critical
+      rules:
+      - metric: mysql_cluster_connected
+        relational_operator: '=='
+        threshold: 0
+        window: 30
+        periods: 1
+        function: min
+    mysql_node_ready:
+      description: "The MySQL service isn't ready to serve queries"
+      severity: critical
+      rules:
+      - metric: mysql_cluster_ready
+        relational_operator: '=='
+        threshold: 0
+        window: 30
+        periods: 1
+        function: min
+  alarm:
+    mysql_check:
+      alerting: enabled
+      triggers:
+      - mysql_check
+      dimension:
+        service: mysql
+    mysql_node_status:
+      alerting: enabled
+      triggers:
+      - mysql_node_connected
+      - mysql_node_ready
+      dimension:
+        service: mysql
+aggregator:
+  alarm_cluster:
+    mysql:
+      policy: majority_of_members
+      alerting: enabled_with_notification
+      match:
+        service: mysql
+      group_by: hostname
+      members:
+      - mysql_check
+      - mysql_node_status
+      dimension:
+        cluster_name: mysql
+        nagios_host: 00-top-clusters

metadata/service/master/cluster.yml

@@ -1,5 +1,7 @@
 applications:
 - galera
+classes:
+- service.galera.support
 parameters:
   _param:
     galera_server_cluster_name: galeracluster
@@ -32,4 +34,4 @@ parameters:
       - name: haproxy
         host: '%'
       - name: haproxy
-        host: ${_param:cluster_local_address}
+        host: ${_param:cluster_local_address}

metadata/service/slave/cluster.yml

@@ -1,5 +1,7 @@
 applications:
 - galera
+classes:
+- service.galera.support
 parameters:
   _param:
     galera_server_cluster_name: galeracluster
@@ -32,4 +34,4 @@ parameters:
       - name: haproxy
         host: '%'
       - name: haproxy
-        host: ${_param:cluster_local_address}
+        host: ${_param:cluster_local_address}

metadata/service/support.yml

@@ -2,12 +2,14 @@ parameters:
   galera:
     _support:
       collectd:
-        enabled: false
+        enabled: true
       heka:
-        enabled: false
+        enabled: true
       sensu:
         enabled: true
       sphinx:
         enabled: true
       config:
-        enabled: true
+        enabled: true
+      grafana:
+        enabled: true