monitoring applications, uptime, log files, etc

Question

How do you monitor your application in production? logs, uptime, etc... (I would prefer an external application, free and open source)

For example, I would like

1. ability to send out alert if the application goes down
2. send alert if cpu usage > than a set threshold
3. send alert if memory usage > than a set threshold
4. send alert for error messages
5. must be configurable, maybe some errors send alert if occurs X times in Y time period

Answer 1

Google Analytics???

Answer 2

What kind of application?

I've used Nagios in the past. It's free and open source. It allows you to setup alerts, monitor event logs, monitor application specific logs, as well as monitor the server infrastructure and network itself.

http://www.nagios.org/

Answer 3

Nagios is the way to go -- a bit of a learning curve, but customizable and powerful. Also has a server-side daemon which can monitor files, disk space etc.

Answer 4

We have a custom in house built piece of monitoring software.

It monitors the event logs on our various live machines (and test) for errors produced by our web applications. All our web applications write any exceptions the error log. It also pings the servers and monitors drive space.

There is a client application on every dev machine that polls the server app that monitors all the servers we have defined. This client app runs in the task tray and pops up messages when anything is out of the norm so a dev sees it instantly. We can also see when testers come across errors and usually have a fix or at least a fix in progress by the time the tester even reports the error.

The server also emails out to a distribution group so that we can see important errors while not at work if we need.

It also has the ability to supress predefined exceptions / errors as well.