Chapter 9. Connecting MOM 2005
To this point, we’ve limited the MOM 2005 discussion to the smallest self-contained unit, the management group. In the preproduction and production management group architecture, the two management groups are independent peers of each other. If one became nonfunctional, the other would continue to chug along none the wiser.
Likewise, all discovery and operational data is restricted to the management group. Reporting data, in the form of published reports, can be consumed outside of the reporting console in a variety of formats, but the data itself remains contained in the MOM 2005 Reporting solution. Basically, all MOM data originates within the management group from the agents and ultimately finds its way into the bit bucket via grooming operations, never having left the boundaries of the management group.
This architecture is a great solution as long as your environment isn’t too big for a single management group, you aren’t required to monitor platforms other than Windows-based servers, or you don’t have administrative and security boundaries between you and untrusted groups. Under these conditions and others, the peer management group architecture won’t meet your needs. Your MOM architecture may need to scale to multiple management groups to monitor many thousands of agents and still provide a single console for resolving alerts. You may need to collect data from or send data to other management tools like those for Unix or mid-range and mainframe ...