4 ms·
so I found it really hard to write the alerts and it has a crappy UI, how's your experience with nagios? what's the fine-tuned host parenting and service depen
by talboren 4y ago
so I found it really hard to write the alerts and it has a crappy UI, how's your experience with nagios?
what's the fine-tuned host parenting and service dependencies exactly?
- polski-g 4y agoUse Thruk as a frontend to Nagios. The nagios web-cgi sucks, indeed. You setup servicedepenency objects so you don't get an alert if multiple services go down when they are interconnected. Let's say you have a service check for for verifying that a web app responds HTTP 200 and another check to confirm postgres is running. If postgres is not running, the web app will not ever respond 200, it would respond HTTP 503 or something. So you create a servicedependency such that the web app depends on postgres, then you will only get one alert instead of two. Host parenting is similar. Lets say you have 2 remote datacenters you are checking in nagios and they are connected in a chain of tunnels. You would set the nearest DC as a parent of the farthest one. Then you would not get ANY notifications for failed checks on farthest DC if you cannot even reach the near DC.
- talboren 4y agothanks for clarifying host parenting and service dependency. gotta say that Thruk looks awful and hard to maintain as well (didn't try it though)