Zabbix extension for monitoring Elasticsearch cluster and node health/status.

Feature:

• jvm stats
• Cluster health:
  • cluster_name
  • status
  • timed_out
  • number_of_nodes
  • number_of_data_nodes
  • total indices docs count
  • total indices deleted docs count
  • primary indices docs count
  • primary indices deleted docs count
  • total indices store size
  • primary indices store size
  • active_primary_shards
  • active_shards
  • relocating_shards
  • initializing_shards
  • unassigned_shards
  • delayed_unassigned_shards
  • number_of_pending_tasks
  • number_of_in_flight_fetch
  • task_max_waiting_in_queue_millis
  • active_shards_percent_as_number

Manual build

git clone https://github.com/zarplata/zabbix-agent-extension-elasticsearch.git

cd zabbix-agent-extension-elasticsearch

make

make install

By default, binary installs into /usr/bin/ and zabbix config in /etc/zabbix/zabbix_agentd.conf.d/ but, you may manually copy binary to your executable path and zabbix config to specific include directory.

Import template_elasticsearch_service.xml and link with host.

WARNING: You must define macro with name - {$ZABBIX_SERVER_IP} in global or local (template) scope with IP address of  zabbix server.

On one node of cluster set MACRO {$GROUPNAME} = REAL_ZABBIX_GROUP. This group must include all nodes of the cluster.

Only this one node will be triggered cluster status (low level discovery added aggregate checks of cluster health).