One of my active projects at work includes a monitoring server for one of our larger companies. With the use of Nagios (Installed with GroundWorkOpenSource) and the NRPE plugin architecture I am able to monitor more things that I can point a stick at.
As a baseline I monitor the utilization of the processor, memory, disks, and auto started services. I have just recently started working with the plugins for Active Directory Domain Controllers, Active Directory Member Servers and hardware monitoring of Dell servers. All of these and more are available from Nagios Exchange as free downloads. The processor, memory and disk monitoring plugins I use are all perl scripts that have been rolled up into exe’s for the masses and work very quickly without perl being installed on the system; the auto started services plugin I use is a vbs script that uses WMI to pull services that are set for auto start on boot and returns a value if something is wrong with any one of them. The Active Directory plugins I use are vbs scripts that use “dcdiag” and “netdiag” from the Windows Server 2003 Support Tools installation. The Dell hardware monitor uses the “omreport.exe” program to check any Dell Poweredge server for hardware related issues. It checks all the same stuff that you would see in OpenManage.
I encourage you to give Nagios a try with or without the use of GroundWorkOpenSource. I use it because it takes only a few minutes and comes preinstalled with a very easy to use GUI for building Nagios configurations and makes it easy to manage lots of hosts.
My next project related to this is to put what I currently have into a managed services model and start offering monitoring as a per host / per month service for my company. Nagios has the ability to be a distributed monitoring solution perfect for this type of setup. I am planning to use OpenVPN as the bridge across networks so that I would have full access to the remote host without opening up any ports on the firewall.