[英]How to configure graphite to track multiple server
We are currently using graphite inside a server farm and we want to track our java apps for all our farm but for each server too. 我们目前在服务器场内使用石墨,我们希望跟踪所有服务器场(但也包括每个服务器)的Java应用程序。
So far it seem that only the global one is working and even if in each apps server we created a different agent with the name of the server. 到目前为止,似乎只有全局服务器正在运行,即使在每个应用服务器中,我们使用服务器名称创建了一个不同的代理。 (they are all different) (它们都不同)
we send the graphite data to a EC2 instance (external to were our instance are located) where graphite is running but the EC2 instance seem too only have the agent of the last loaded server. 我们将石墨数据发送到正在运行石墨的EC2实例(在我们实例所在的外部)中,但EC2实例似乎也只有最后加载的服务器的代理。
Can I be that since our request are going through a proxy graphite consider that they are from the same source even if we give them different agent name ? 因为我们的请求正在通过代理石墨进行处理,所以即使我给他们提供不同的代理人名称,我是否也可以认为它们来自同一来源?
How can we prevent from logging all metric under one agent ? 如何防止在一个代理下记录所有指标?
The only solution I see right now is to put graphite withing the same network the farm is but that would take week since we have to wait for another group to give us acces and configure other box. 我现在看到的唯一解决方案是将石墨置于场中相同的网络中,但这将需要数周的时间,因为我们必须等待另一组为我们提供配件并配置其他盒子。
Why it wasn't working it's because we were using dns instead of the ip. 为什么它不起作用,是因为我们使用的是dns而不是ip。 Use the ip of your instance and it shouldn't be a problem. 使用您实例的ip,这应该不是问题。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.