Metric alerts monitor data that is very dynamic in nature and attempt to evaluate
how the system is performing. For example, is performance acceptable, is the
target available, and so on. Typically when an alert does occur, the administrator
should address the problem as soon as possible in order to get the system back
in a usable state.
Policies, on the other hand, monitor data that changes much less frequently
and evaluate the existing state of the system against corporate standards and
best practices. In some cases, the resolution of a policy violation may take
as much as a number of days to completely implement and may even involve some
coordinated effort. For example, administrators may not be able to merely rollout
critical patches to production targets without:
Testing them first in a development environment
Getting appropriate approval
Timing the actual patch application to the production target during some scheduled maintenance window for that target
In some companies, alerts and policies are seen as serving very different purposes.
For example, they require certain policies to be monitored and enforced; that
is, the policy monitoring is a corporate-level requirement. If administrators
want to disable a policy, they must submit a formal request along with a justification
as to why.
On the other hand, these companies may have no formal requirements that alerts
be monitored. If administrators decide not to monitor alerts, that is their
prerogative. Eventually, if there is a problem, the administrators will have
to deal with irate end users once availability and performances issues arise.
Enterprise Manager Information Roadmap
Copyright © 1996, 2007, Oracle. All rights reserved.