Zabbix监控es集群状态-运维开发故事

不患人之不己知，患其不能也

《论语》

es提供了一个可以获取集群健康状态的api，访问http://esurl:9200/_cluster/health?pretty
和 Elasticsearch 里其他 API 一样，cluster-health 会返回一个 JSON 响应。

响应的内容解释：

“cluster_name”: “my-es”, #集群名
“status”: “yellow”, #集群健康状态，正常的话是green，缺少副本分片为yellow，缺少主分片为red
“timed_out”: false,
“number_of_nodes”: 1,#集群节点数
“number_of_data_nodes”: 1,#数据节点数
“active_primary_shards”: 15,#主分片数
“active_shards”: 15,#可用的分片数
“relocating_shards”: 0,#正在迁移的分片数
“initializing_shards”: 0,#正在初始化的分片数
“unassigned_shards”: 15, #未分配的分片，但在集群中存在
“delayed_unassigned_shards”: 0, #延时待分配到具体节点上的分片数
“number_of_pending_tasks”: 0, #待处理的任务数，指主节点创建索引并分配shards等任务
“number_of_in_flight_fetch”: 0,
“task_max_waiting_in_queue_millis”: 0,
“active_shards_percent_as_number”: 50 #可用分片数占总分片的比例

1.编写采集脚本获取集群状态

vim /monitor_es.py

#encoding=utf-8import requests,jsonimport sys
headers={"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3626.109 Safari/537.36"}
response=requests.get("http://192.168.179.133:9200/_cluster/health",headers=headers)
s=json.loads(response.content.decode())
parm=sys.argv[1]

itemlist=["cluster_name","status","timed_out","number_of_nodes","number_of_data_nodes","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"]if parm not in itemlist:
    print("parm failed")
    sys.exit(1)else:
    print(s[parm])

给脚本赋予执行权限
chmod +x /monitor_es.py

2.修改zabbix-agent配置文件

vim /etc/zabbix/zabbix_agentd.d/userparameter_mysql.conf
UserParameter=es.[*],/usr/bin/python /monitor_es.py $1
重启zabbix-agent
sytemctl restart zabbix-agent