bjoernvold.com

I have successfully installed Nagios and integrated it in my LAMP server (you do not need a server). I created the user and copied the config files to the right places.
Nagios is supposed to be the best open source host, service and network monitoring program. It monitors specified hosts and services, alerting you to any developing issues, errors or improvements. I have also heard that it can be used as a server for thin clients without straining the server becaue more of the process is happening on the client side.

Well that sounds good, so I had to try it of course.
So here I am after a successfull login:

Now I only have to start using it I'll be back with the dramatic stroy.....
Meanwhile if you want to install it; the Fedora Quickstart has the most global description (file locations etc).
I belive distros like Mandriva and Mageia have a task-nagios that will do the hard work for you and place things correctly.
I placed everything myself which has the advantage that I know where it went.

And this thing is practical indeed. Nagios is a framework and will not do much without plugins. But the plugins are downloadable as a package (and part of most meta packages).
So I was pretty much set after the successful install, and started monitoring the server.
And the big difference is that nagios keeps you informed via your browser and by email or whatever you choose. Being so flexible you can completely eliminate the "Windows syndrome" where you get cryptic and non significant messages all the time. But there is not much to do it is pretty much set up like that already when you use the default plugins.

So if a server is getting overloaded you will see exactly when on the reports at http://local(or some)host/nagios for diagnostic reasons this is great and bit different from the live surveillance tools that I am used to.

Nagio is both and it keeps the logs in var if you like to fig for them there but why should you when Nagios gives you such beautiful reports? Right there on the web page!

It is understandable that it has become the "industry standard" and it's main strength is the flexibility - that makes it a favorite in my eyes.
You have the trends for a period (default is last 7 days)

Or the most recent allerts:

(Don't worry this is a test, so I had to test it )

You will probably have to read some docs and tweak a little before you have it all up and running but it is worth it

Looks like a good tool for mega website admins, like bjoernvold. I'll keep Nagios in mind for when lanlesalon.com goes global.

Yes this is a tool for mega stuff. But it can be used to monitor you local Puppy server too. But I must admit that I have integrated it in an Apache server. If you service others you can watch the server history via the internet and identify the bottlenecks.
You could have a look at the running services at the local hair sallong or at the Siemens head office.
And you can set Nagios to autofix stuff - not just report it.

Some updates later this just stopped working

Whoops!

Error: Could not read host and service status information

!

So forget all the "standard" talk above. The web is full of questions regarding this bug - and no answers.
http://bit.ly/YNMkLg
So it is reduced to crap now! Removing it......