On Sun, Feb 8, 2015, at 19:36, Andrew Rakowski wrote:
I've been using Big Brother since 1999, and Xymon for the last couple of years (on a different project at the lab), but recently, a team member has suggested that we switch infrastructure monitoring to Nagios, which he's been using on other systems he manages elsewhere in the lab.
Why? What is broken or deficient with your current Xymon implementation? If there isn't an answer other than familiarity you're going down a dangerous path. Is your existing monitoring reliable? If so, then keep it. Changing monitoring systems and not knowing how it will behave is not fun.
Xymon tends to win for me because I don't run large homogenous systems. If I had 3000 servers that all had identical services being monitored Nagios will work pretty well -- create your rulesets, put all 3000 servers in the group.
If you have a lot of heterogeneous systems to monitor Xymon will win every time in its simplicity for setup and maintainability. Plus you don't need to configure the client any further than telling it what the Xymon server's IP is, unless you're adding additional monitoring scripts.
It doesn't matter what you wrap Nagios in, it's still Nagios underneath and has the same problems. You're just expecting yet another layer of complexity to solve them for you.