Chapter 1. Introduction to business systems management 11
Filtering is a powerful feature for building a BSV by providing ad-hoc selection
criteria, such as object type, name, and alert state. This enables the command
center staff to quickly create a custom view to closely monitor a collection of
objects showing recent trouble conditions. Filtering also allows representations of
the same object contained in different BSVs to filter events differently, therefore
allowing you to be notified only on events that pertain to you.
IBM Tivoli Business Systems Manager monitors resources for state changes and
the performance characteristics that reflect their availability. These resources are
represented by IBM Tivoli Business Systems Manager objects in the SQL
database. Actions on an object, such as an alert notification and the propagation
of that alert up and down a view, result from events. Events may be exceptions
associated with an object or a state change of that object. Exceptions occur
when the counters that measure performance thresholds are exceeded. An
example of an exception could be unacceptable response time associated with a
CICS transaction. Another example of an event could be the receipt of a console
message that a batch job terminated abnormally. This would cause a state
change to occur and would result in an event. As events occur within the
monitored environment, they are collected and recorded by IBM Tivoli Business
Systems Manager, and are displayed by tagging an alert icon on the offending
object’s icon.
Propagation leverages the object-orientated implementation inherent within the
IBM Tivoli Business Systems Manager environment and continuously
disseminates events throughout the object hierarchy. Propagation escalates
alerts up the hierarchy based on the severity of events and the volume and rate
at which they occur. Exceptions, console messages, and other events are
assigned priorities for each object. When an object receives an event, the events
priority is examined and compared against tolerance rates set for that object. If a
threshold is exceeded, an alert occurs on that object and sends an event to its
parent object on the hierarchy. This, in turn, can cause another event to occur
and another alert to be sent further up the hierarchy. In addition to controls that
adjust rates for incoming performance exceptions, each object on the hierarchy
includes controls for events arriving from the child objects below it.
Figure 1-3 on page 12 shows an alert occurring on the DB2 subsystem D7Q2
object under the SC69 system. The event is propagated up the hierarchy to the
Enterprise level. The propagation also takes place on the Business System View
of the object that affects the ITSO RESOURCES object.