Question

I need to monitor if Graphite is loosing metrics. I already use https://github.com/klen/graphite-beacon to send alerts on e-mail everytime when executing the query from the following json and the values are breaking the rule

{
  "interval": "5minute",
  "logging": "debug",
  "host": "servmail01",
  "port": 25,
  "from": "BeaconAlerts@metrics01.local",
  "to": ["myemailaddress@email.com"]
  },

  "alerts": [
      {
      "name": "fullQueueDrops",
      "query": "transformNull(carbon.relays.metrics01-a.destinations.127_0_0_1:2104:a.fullQueueDrops,0)",
      "method": "average",
      "interval": "5minute",
      "rules": [ "critical: > 1" ]
    }
  ]
}

The thing is that I would like to monitor the same with zabbix (that does any other monitoring stuff on our network and I would love zabbix to send e-mails when I have metrics dropped. So I have looked to see if there's any tool available and I got into https://github.com/blacked/graphite-to-zabbix. The only problem with it is that it lacks a better documentation.

Has anyone used it? I understood that I should create a template for metrics in zabbix (which I did), to create an item corresponding to the metrics I want to send, but at some point in the documentation it says Note that g2zproxy will work with zabbix web api specified in -z argument, but it will send metrics to service specified in /etc/zabbix/zabbix_agentd.conf. What and how do I specify in that config the service? I don't get it.

Thank you. Gabriel

Answer 1

所以最后我明白了为了使这项工作必须做些什么。这并不难。

在zabbix中我创建了一个项目

graphite[transformNull(carbon.relays.metrics01-a.destinations.127_0_0_1:2*04:*.fullQueueDrops,0)]

完全取自石墨图数据。

cron作业通过将数据发送到zabbix来工作，并且创建的项目选择数据并存储它。根据此项目，您可以创建图表。

这就是全部。我希望任何使用它的人都能得到健康。我要感谢Alexey Dubkov的出色工作。 Alexey's github project

加布里埃尔

Monitoring Graphite fullQueueDrops

1 个答案: