generalize the concept of a device template

Today, the OpManager's concept of device template is very limiting. It doesn't include service monitors, url monitors, process monitors, and script monitors. Furthermore, I cannot use one device template for one bank of linux based web servers and a different device template for a different bank of linux based web services servers, and so on and so forth.

A large server farm typically have different groups of machine for different purpose. Each machine in a particular group is similarly configured and are running the same set of services and software application. A better device template concept is as follow:

1. Allow arbitrary grouping of machines base on purpose (not just snmp sysOID).

2. Allow definition of device template for a machine group

3. Each device template should support all the different monitor types (service monitor, url monitor, performance monitor, process monitors and script monitor)

The key to having a good monitoring setup is the enforcement of consistency of monitoring. This is particularly important for large production environments (a must have). A better device template mechanism will be a huge improvement in this area for OpManager.