Enterprise Manager 13 Monitoring Strategies

Strategies for Scalable,
Smarter Monitoring using

Oracle Enterprise Manager
Cloud Control 13c
April 2020 | Version 1.00

Copyright © 2020, Oracle and/or its affiliates
PURPOSE STATEMENT
This document provides an overview of features and enhancements included in Oracle Enterprise Manager 13c. It is
intended solely to help you assess the business benefits of upgrading to Oracle Enterprise Manager 13c and to plan your I.T.
projects.
DISCLAIMER
This document in any form, software or printed matter, contains proprietary information that is the exclusive property of
Oracle. Your access to and use of this confidential material is subject to the terms and conditions of your Oracle software
license and service agreement, which has been executed and with which you agree to comply. This document and
information contained herein may not be disclosed, copied, reproduced or distributed to anyone outside Oracle without
prior written consent of Oracle. This document is not part of your license agreement nor can it be incorporated into any
contractual agreement with Oracle or its subsidiaries or affiliates.
This document is for informational purposes only and is intended solely to assist you in planning for the implementation
and upgrade of the product features described. It is not a commitment to deliver any material, code, or functionality, and
should not be relied upon in making purchasing decisions. The development, release, and timing of any features or
functionality described in this document remains at the sole discretion of Oracle.
Due to the nature of the product architecture, it may not be possible to safely include all features described in this document
without risking significant destabilization of the code.
1 WHITE PAPER | Strategies for Scalable, Smarter Monitoring Using Oracle Enterprise Manager Cloud Control 13c | Version 1.00
Copyright © 2020, Oracle and/or its affiliates | Public
TABLE OF CONTENTS
Purpose Statement 1
Disclaimer 1
Product Overview 3
Introduction 3
Laying the Groundwork for Monitoring 3
Define and Set Up Standards for Monitoring 4
Organize Targets Based on How They are Monitored 4
Defining Your Administration Group Hierarchy 6
Setting Target Property Values 8
Define Standards for Monitoring 10
Associate Template Collections with Administration Groups 11
Set the Synchronization Schedule 14
Define Roles for Different Job Responsibilities 15
Plan Job Responsibilities (Who can do what operations?) 15
Understand Enterprise Manager Privileges to support job responsibilities 16
Use roles to manage user privileges 18
Leverage the privilege-propagating nature of Administration Groups 19
Create Roles for different job responsibilities 19
Assign Roles to your Enterprise Manager administrators 21
Set up Rules to Manage Events and Incidents 21
Events and Incidents 21
Understand Rule Sets and Rules 23
Use Groups in Rule Sets 23
Auto-create Incidents Using Rules 24
Minimizing Event Storms by Compressing Multiple Events into a Single Incident 24
Plan Your Rules 24
Which type of rule should I use? 26
Other Considerations for Planning Rules 29
Implement the Rules in Enterprise Manager 32
Benefitting From Economies of Scale: A Fully Automated and Scalable Monitoring Setup 32
Managing Incidents 33
Additional Monitoring Requirements and Recommendations 35
Auto-fixing Alerts using Corrective Actions 35
Monitoring When Enterprise Manager is Under Planned Maintenance 36
Sharing the Administration Group Hierarchy across Different Teams 36
Using Administration Groups for Other Group Operations 38
Verifying Targets are Part of the Administration Group Hierarchy 40
Changing the Administration Group Hierarchy after Initial Creation 41
Verifying Targets are in Sync with Your Monitoring Standards 44
Enabling Events for Jobs 45
Integrating with Third Party Event Systems and Service Desks 45
Too Many Alerts 45
Too Many Emails When a Host Goes Down 47
Using Root Cause Analysis for Target Down Events 49
Managing Diagnostic Incidents and Problems 52
Conclusion 53
PRODUCT OVERVIEW
Oracle Enterprise Manager is Oracle’s integrated enterprise IT management product line and provides
the industry’s first complete cloud lifecycle management solution. Oracle Enterprise Manager’s
Business-Driven IT Management capabilities allow you to quickly set up, manage and support
enterprise clouds and traditional Oracle IT environments from applications to disk. Enterprise Manager
allows customers to achieve:
• Best service levels for traditional and cloud applications through management from a business
perspective
• Maximum return on IT management investment through the best solutions for intelligent
management of the Oracle stack and engineered systems
• Unmatched customer support experience through real-time integration of Oracle’s knowledge
base with each customer environment
INTRODUCTION
Whether you’re supporting enterprise clouds or traditional IT applications, the need for the proactive
and complete monitoring of your business applications and their underlying IT infrastructure, on-
premises or in the cloud, continues to be a critical requirement for any datacenter. Effective monitoring
of today’s rapidly changing datacenter requires a management tool that can scale dynamically as the
enterprise grows and IT staff that can use the tool in conjunction with best practice standards and
processes. Enterprise monitoring has always been built into Enterprise Manager’s DNA since its initial
release. Over time it has evolved to work seamlessly with new Oracle technologies and meet the ever-
changing requirements of IT staff who use it to manage the Oracle footprint in their datacenters. In the
years that we’ve worked with customers in their Enterprise Manager deployments, we’ve sought to
understand their product requirements, processes and strategies used to monitor their own
datacenters. This white paper is a consolidation of these monitoring best practice strategies used in
conjunction with product capabilities in Enterprise Manager 13c. These strategies are meant to provide
high level guidance in using Enterprise Manager’s monitoring features to:
• Meet your datacenter’s monitoring requirements
• Comply with security best practices
• Provide a solution that is easy to set up and manage
• Provide a solution that scales as your enterprise grows
These monitoring strategies are divided into three sections:
• Laying the Groundwork for Monitoring
• Managing Incidents
• Additional Monitoring Requirements and Recommendations
LAYING THE GROUNDWORK FOR MONITORING

Setting up your enterprise for monitoring requires a combination of planning and implementation. In
order to manage and monitor any large datacenter to scale while meeting SLAs and other business-
related requirements, it is important to have some uniformity in the way managed entities (called
targets) are monitored and standard IT management procedures to handle events and incidents raised
on those targets. At a minimum, there are three major areas to address when it comes to setting up
target monitoring:
• Defining and setting up a standard set of monitoring settings for your targets
• Defining and granting the appropriate level of privileges in Enterprise Manager to
administrators who are responsible for managing these targets
• Setting up rules to automate IT operational processes such as sending email notifications for
events, opening helpdesk tickets, escalating long running events, etc.
For each area, the IT staff has to determine the specific requirements for their datacenter based on
business needs. For example, they will need to determine the appropriate set of metrics and thresholds
to monitor their production servers in order to meet SLAs. They will need to determine which
administrators are responsible for operations such as defining monitoring settings, responding to
events, etc. and then grant them the appropriate privileges in Enterprise Manager in support of those
operations. It is important to take time to plan and define these in advance before anything is
implemented in Enterprise Manager. Assuming such planning has already taken place, this paper will
recommend strategies for implementing these plans in Enterprise Manager in a scalable way. This
means you will only need to set it up once, and as your enterprise grows and more targets and/or
administrators are added, the additional setup needed to accommodate the growth will be kept to a
minimum because your monitoring setup will be automatically leveraged. Each of these areas will now
be discussed in more detail.
Define and Set Up Standards for Monitoring

This first phase involves defining how your targets should be monitored in Enterprise Manager. Several
steps are involved and each of these steps is discussed in detail below.
Organize Targets Based on How They are Monitored

Not all targets are alike. Some support mission critical applications, some support test environments
and others support development environments. Based on these different usage profiles, most
datacenters are likely to have different sets of monitoring settings (i.e. metrics, thresholds, collection
schedules, corrective actions) for each of these different usage profiles. (The details of specific metrics
to monitor, the threshold values to use, etc. are outside the scope of this paper but should be
determined by the appropriate administrators in your IT organization). For example, mission critical
targets might be monitored more comprehensively for availability, performance and space usage while
test targets might only be monitored for basic availability. It is first important to identify these different
monitoring profiles in your datacenter and then plan on how you can group your targets based on
these monitoring profiles.
As a simple example, you might have this type of grouping:
Figure 1. Monitored targets are divided into Production targets and Non-Production targets
In Figure 1 above, the set of targets monitored in Enterprise Manager are divided into 2 groups: all
production targets in one group because they are all monitored using production monitoring settings
and all non-production targets such as test and development targets are put in another group because
they are all monitored using non-production settings.
Suppose there are additional monitoring requirements within the group of production targets. For
example, the targets supporting the production applications in the Sales Line of Business might have, in
addition to the production monitoring settings, additional metrics that need to be monitored. The
targets supporting the applications in the Finance Line of Business might also have additional metrics
and/or metric thresholds that are slightly stricter than the thresholds used for the other production
targets. This scenario leads to this type of grouping:
Figure 1. Monitored targets further broken down by Line of Business
In Figure 2 above, you will notice there is a further breakdown of the Production Targets group into
different subgroups based on Line of Business. The grouping at this level is again based on how
targets are monitored, where targets monitored in the same way are put together in the same group.
So, under the Production targets group, the Sales targets have additional monitoring settings, the
Finance group has additional monitoring settings and all other Lines of Business (HCM, Manufacturing,
and Marketing) that don’t have these additional monitoring settings are put together in another group.
A similar scenario can happen for the targets in the Non-Production group.
When designing your Administration Group hierarchy, it is important to remember that the primary
goal here is to define the group hierarchy based on how targets are monitored. While you can re-use
the same hierarchy for other purposes (e.g. reporting, job submission), designing the hierarchy with
these other applications in mind might result in a hierarchy that is bigger and more complicated. You
will need to weigh the benefits that such a hierarchy may provide versus the cost of managing any
added complexity.
Once this grouping has been defined, it is important to understand how this group hierarchy is
specified and implemented in Enterprise Manager. In Enterprise Manager, each node in the group
hierarchy is a (target) group and each level in the group hierarchy is identified by a target property. (A
target property is an attribute associated with all targets in Enterprise Manager and is used to annotate
operational characteristics of the target, e.g. Line of Business, Owner, etc.). Within a level in the group
hierarchy, specific values of the target property determine the membership criteria of groups at that
level. In Figure 3 below, the target property Lifecycle Status is used to identify the two groups’
membership criteria The membership criterion for the Production Targets group is that the target’s
Lifecycle Status property should be ‘Production.’ The membership criteria of the Non-Production
Targets group (consisting of test targets, development targets or staging targets) is that the target’s
Lifecycle Status property should be Development or Test or Staging. A target cannot be added to these
groups directly, rather, its target properties need to be set such that it matches the group’s membership
criteria. Once that happens, Enterprise Manager will automatically add the target to the appropriate
group. For example, if you set a target’s Lifecycle Status property to ‘Production’, it will automatically
be added by Enterprise Manager to the Production Targets group. These types of groups are called
Administration Groups and together they form an Administration Group hierarchy.
Figure 2. In this Administration Group hierarchy, the Lifecycle Status target property is used to define the membership criteria
of the 2 groups.
Defining Your Administration Group Hierarchy

Administration Groups are a special type of group designed primarily to deploy monitoring settings to
targets as they join the group1. (Details of how this occurs will be discussed later.) Defining your
Administration Group hierarchy involves defining the target properties that make up each level of the
hierarchy and the values of the target properties that determine the different groups at each level. In
Figure 4 below, the first level of Administration Group is based on Lifecycle Status target property and
the second level is based on Line of Business. The membership criteria for the Prod-Sales group are
Lifecycle Status = Production and Line of Business = Sales. The membership criteria for Prod-Finance
are Lifecycle Status = Production and Line of Business = Finance. The membership criteria for Prod-
Others is Lifecycle Status = Production and Line of Business = HCM or Manufacturing or Marketing. You
will need to specify the exact list of values for each target property used as membership criteria. In
Enterprise Manager, the creation of the Administration Group hierarchy involves choosing the target
property that defines the level and for that level, specifying the values of the target properties that
define each group for that level. For a complete list of target properties supported for Administration
Groups and for more details on creating the Administration Group hierarchy, refer to the
“Implementing Administration Groups and Template Collections” chapter of the Oracle Enterprise
1You can also use Administration Groups to automatically deploy Compliance Standards and Cloud
Policies to targets as they join the group. However, that is outside the scope of this paper.
Manager Cloud Control Administrator’s Guide 13c. Figure 4 below shows the creation of the
Administration Group hierarchy in Enterprise Manager.
Figure 3. In this Administration Group hierarchy, the Lifecycle Status and Line of Business target properties define the
membership criteria of the groups.
Regardless of the number of levels you define in your Administration Group hierarchy, it is important to
note three points:
• Targets are always added to the leaf-level groups of the hierarchy.
• A target in the hierarchy can directly belong to at most one Administration Group. (It can
directly belong to any number of regular, non-Administration Groups). As you’ll see in the
later sections of this paper, this is to prevent the potential ambiguity that can arise from
different monitoring settings if a target is part of multiple Administration Groups associated
with different monitoring templates.
• A target must match all membership criteria defined by the levels of the hierarchy in order to
join an Administration Group. For example, in the Administration Group hierarchy in Figure 4,
a target’s properties must be set for both Lifecycle Status and Line of Business. In addition, the
values set for these properties must match the membership criteria of one of the
Administration Groups. If a target only has its Lifecycle Status property set, it will not join the
Administration Group because its Line of Business has not been set. If a target has Lifecycle
Status=Production and Line of Business=IT, then it will not join the Administration Group
because there is no Administration Group with that membership criteria.
Figure 5. When creating the Administration Group hierarchy, specify the target property for each level and the values of the
target property that define each group in that level.
Setting Target Property Values

As mentioned, for a target to join an Administration Group, its target properties need to be set such
that it matches the membership criteria of the Administration Group.
You can set the value of a target’s properties as part of its process of being added to Enterprise
Manager. If the target is added via the console using either the manual target addition or target
promotion workflow, there are steps in the workflow that enable the user to specify the target
properties of the target that is being added.
Enterprise Manager administrator accounts also have attributes. These include: Contact, Location,
Department, Cost Center, and Line of Business. If you set values to any of these attributes, then any
target added by the administrator will automatically have the same values set to its target properties.
For example, if the administrator’s Line of Business is set to Finance, any targets added by that
administrator will have its Line of Business also set to Finance by default (this can always be
overwritten). Hence if any of these properties are used as Administration Group criteria, to facilitate
setting of target properties, consider setting the values at the administrator level and use that
administrator account to add targets to Enterprise Manager.
If the target is added using Enterprise Manager Command Line Interface (EM CLI2) add_target verb and
the attributes of the administrator who added the target was not set or did not match the
Administration Group criteria, you can follow that with another verb set_target_property_value to
specify the values of the target properties. In the following example, this EM CLI command sets the
Lifecycle Status and Line of Business properties for a database target that has been added to Enterprise
Manager:
$ emcli set_target_property_value
-property_records="MyDB:oracle_database:LifeCycle
Status:MissionCritical;MyDB:oracle_database:Line of Business:Finance"
2EM CLI is Enterprise Manager’s command line utility that enables you access to Enterprise Manager
functionality within scripts.
The EM CLI verb set_target_property_value is the recommended way to set target properties in bulk
across many different targets at a time. For more details on the EM CLI verbs, refer to the Oracle
Enterprise Manager Command Line Interface documentation.
To specify target properties of a target via the console, go to the target’s Target Properties page which
is accessible via Target menu → Target Setup → Properties. The Target Properties page that appears
will allow you to set properties for the target. As a reference, it will also display the membership criteria
of the Administration Group hierarchy.
Figure 6. The Edit Target Properties page shows you the criteria used for the Administration Group hierarchy.
The All Targets page (accessible from Targets menu → All Targets) allows you to select additional
columns for the All Targets table. The columns that you can add to the table include target properties.
Figure 7. The All Targets page allows you to select target properties as additional columns to be displayed in the targets
table.
If you customize the All Targets page by adding the target properties used as Administration Group
criteria, you can easily see the target properties defined for the targets and also find out which targets
do not have their target properties set. Note that a maximum of 2000 targets are shown in this page. If
you’re looking for a target and it’s not in the list, you can use the Search Target Name field to filter the
list to show targets matching the name specified. To set a target’s properties, select the target from the
table, right-mouse click to open up the target menu, and select Target Setup → Properties to access the
Target Properties page.
One important note about setting target property values for aggregate targets, i.e. targets that contain
other targets (e.g. Oracle WebLogic Domain target containing Oracle WebLogic Servers): if the intent is
to set the target properties for the aggregate target itself as well as its member targets, then you will
need to use the EM CLI verb set_target_property_value on the aggregate target and use the option –
propagate_to_members. This will set the target property values on the aggregate target itself as well as
all of its current member targets. Any new member targets that will be added in the future will NOT
have its target properties automatically set; hence you will need to set the appropriate target property
values for any members added in the future. The reason for not automatically setting the target
properties for members of these aggregates is because these member targets could potentially be part
of other aggregate targets with different values specified for their target properties. Hence, the
administrator must decide and set the appropriate values of target properties for such targets.
The following EM CLI example sets the Location target property of a database system (aggregate
target) and all its members using the –propagate_to_members option:
$ emcli set_target_property_value
-property_records="dbrac_sys:oracle_dbsys:Location:Bangalore"
-propagate_to_members
Cluster targets on the other hand are aggregate targets that have strong membership semantics, i.e.
members of clusters can only belong to one cluster aggregate. Examples of these targets types are
Redundancy System, Database Cluster, Host Cluster, etc. If you set a target property for a cluster target,
then the same target property automatically applies to all members of the cluster target including any
new members that will be added in the future.
Define Standards for Monitoring

The next step is to define the monitoring settings for the groups in the Administration Group hierarchy.
For example, in our Administration Group hierarchy, you’ll need to define the monitoring settings
(metrics and associated thresholds) for the targets in the production group and the set of monitoring
settings for the targets in the non-production group. While Enterprise Manager’s out-of-box
monitoring settings for targets might provide some initial guidance, IT organizations should give some
thought in determining what’s appropriate for their targets. Determining this appropriate set of metrics
and thresholds for targets is outside the scope of this paper, but there some general guidelines
provided in the “Additional Monitoring Requirements and Recommendations” section that is available
in the latter part of this paper (refer to the subsection “Too Many Alerts”). Once you’ve defined these
monitoring settings, create monitoring templates in Enterprise Manager containing these monitoring
settings. (Monitoring templates are named collections of metric settings for a target type. These
include metrics, their associated collection schedules, thresholds, and optionally, corrective actions).
Monitoring templates are defined per target type so create one monitoring template for each of the
target types in your Administration Group. For example, if your Production Targets group contains
hosts, databases and listeners, create three monitoring templates -- one for host targets, another for
database targets and another for listener targets. See Figure 8 on the next page.
Figure 8. Sample host monitoring template
Once you’ve created the monitoring templates, combine these together into a container called a
Template Collection. A Template Collection is a collection of different monitoring templates designed
to specifically contain the monitoring templates for an Administration Group. For our Administration
Group in Figure 3, you will end up defining a Template Collection (containing monitoring templates) for
the Production Targets group and another Template Collection for the Non-Production Targets group.
Refer to the “Implementing Administration Groups and Template Collections” chapter of the Oracle
Enterprise Manager Cloud Control Administrator’s Guide 13c for specific steps in creating Template
Collections.
Note: In all our diagrams, the top-level node of the Administration Group hierarchy is labeled “All
Targets” because of the expectation that all targets in the Enterprise Manager site will be monitored
using the Administration Group – Template Collection setup. However, it is possible that some
datacenters may choose not to include all targets in the Administration Group hierarchy. If this is the
case, then the top-level node really represents the group of all targets that are part of the
Administration Group hierarchy and not all the targets in the Enterprise Manager site. It is a good
practice to periodically check the Unassigned Targets Report (accessible from the Administration Group
UI in the console) to ensure that no target that is meant to be part of the Administration Group
hierarchy has been missed. Such targets could be missed because its target properties have not been
set to match the membership criteria of any Administration Group.
Associate Template Collections with Administration Groups

After you’ve created the Template Collections, the next step is to associate these with the appropriate
Administration Group. In Enterprise Manager, this is done by selecting the Administration Group,
clicking the associate button and choosing the Template Collection (see Figure 9). Refer to the
documentation (“Implementing Administration Groups and Template Collections” chapter of the Oracle
Enterprise Manager Cloud Control Administrator’s Guide 13c) for details on this step. Once this
association is done, then any target added to the Administration Group will automatically be applied
with the monitoring settings from the associated Template Collection.
Figure 9. Associate the PROD Template Collection with the PROD Administration Group.
For the 3 level Administration Group hierarchy in Figure 4, all the groups under the Production Targets
group start off with the production monitoring settings. Hence a Template Collection (e.g. Prod
Template Collection) can be created containing monitoring templates with the production monitoring
settings. This Template Collection should then be associated with the Production Targets group. All
targets in subgroups Prod-Sales and Prod-Others will be applied with the monitoring settings. The
Prod-Finance Group had monitoring settings required in addition to the production monitoring
settings. Create monitoring template(s) containing these additional monitoring settings and put them
in another Template Collection. Associate this new Template Collection with the Prod-Finance Group
(see Figure 10 below). Targets in the Prod-Finance Group will be applied with a union of monitoring
settings from the Prod Template Collection and Prod-Finance Template Collection. If there are any
metrics in common between the two templates, the metric settings from Prod-Finance Template will
take precedence.
Figure 10. Associate the PROD-Finance Template Collection with the Prod-Finance group. This template collection contains
additional templates for the Prod-Finance group.
To see the combined (final) set of monitoring settings that will be applied to targets in the Prod-Finance
group (or any group), select the group and then click on the Show Group Management Settings option.
This is especially useful in scenarios where monitoring settings to be applied to targets are based on
multiple monitoring templates. In the subsequent page that appears, you can review the combined
(aggregate) monitoring settings that will be applied to each target type in the group, check the
synchronization status of members (if any) in the group and quickly go to any of its parent groups to
perform similar actions. Refer to Figure 11 for more details.
Review monitoring
settings that will be
applied to each target
type in the group
Check if targets have

monitoring templates
applied, shown as
‘Synchronized Targets’.
Figure 11. Group Management Settings page enable you to review monitoring settings for the group and check the
synchronization status of its members.
Defining Monitoring Templates for Primary and Standby Databases

When monitoring primary and standby databases, there are a set of metrics and thresholds that apply
to the database based on its current role – primary or standby. If there is a role change, i.e. the primary
becomes the standby and the standby becomes the primary, you want to ensure that the applicable
metrics and thresholds that pertain to the new standby and new primary are correspondingly changed.
For example, the Redo Generation Rate (KB/second) metric applies to the primary database and the
Apply Lag (seconds) metric applies to the standby database. Let’s say you have DB1 (primary database)
and DB2 (standby database) with the following metric settings:
• DB1 (primary database)
o Redo Generation Rate (KB/second) metric: Warning = 25, Critical = 150 thresholds
• DB2 (standby database)
o Apply Lag (seconds) metric: Warning = 7 and Critical = 17 thresholds
If there is a role change and DB1 becomes the standby and DB2 becomes the primary, you would like to
have the following metric settings:
• DB1 (standby database)
o Apply Lag (seconds) metric: Warning = 7 and Critical = 17 thresholds
• DB2 (primary database)
o Redo Generation Rate (KB/second) metric: Warning = 25, Critical = 150 thresholds
To accomplish this, first make sure the database monitoring template contains the union of database
metrics and thresholds that apply to both the primary and standby database. In our example above,
make sure the monitoring template has:
• Redo Generation Rate (KB/second) metric: Warning = 25, Critical = 150 thresholds
• Apply Lag (seconds) metric: Warning = 7 and Critical = 17 thresholds
Set the target properties of both the primary and standby databases so that they join the same
Administration Group. Add the monitoring template to the template collection that will be associated
with the Administration Group (or parent Administration Group) to which the databases belong. When
the template settings are initially applied, the appropriate metrics and thresholds will be applied to the
database based on its role. In our example, DB1 (primary) will have Redo Generation Rate (KB/second)
metric with the corresponding thresholds and DB2 (standby) will have Apply Lag (seconds) metrc with
the corresponding thresholds. If there is a switchover later between the primary and standby, the
corresponding database roles for the databases will be updated (i.e. DB1 becomes the standby and DB2
becomes the primary) and the associated monitoring template will be re-synced with the targets. This
wil cause the Redo Generation Rate (KB/second) metric thresholds to be applied to the new primary
(DB2) and the Apply Lag (seconds) metric thresholds to be applied to the new standby (DB1).
If primary and standby databases need to be in different Administration Groups, ensure that the
monitoring templates associated with their respective Administration Groups contain a union of the
database metrics that apply to primary and standby databases. The monitoring templates associated
with these different Administration Groups can have different threshold values for the database
metrics.3 When the primary and standby databases join their respective Administration Groups, the
appropriate metric threshold settings will be applied based on their current database roles. Later, when
a switchover occurs, the database roles will be updated for the databases and the associated
monitoring templates will be automatically re-synced (i.e. re-applied) with the appropriate metric
thresholds based on their new database roles.
Set the Synchronization Schedule

The synchronization schedule determines when sync operations are executed by Enterprise Manager.
Sync operations refer to the application of monitoring templates within a Template Collection to the
relevant targets in the associated Administration Group. When a target is added to an Administration
Group, then the associated monitoring template is automatically applied to the target. However, there
are other conditions under which templates need to be applied to a target in an Administration Group:
• When a Template Collection containing monitoring templates is initially associated with an
Administration Group that already has member targets or
• When there are changes made to a monitoring template that is part of a Template Collection
associated with an Administration Group or
• When there are changes made to the monitoring settings of a specific target and the option to
prevent ‘Template Override’ has not been specified
In all these cases, the targets in the Administration Group need to be made ‘in sync’ with the associated
monitoring template by applying the monitoring template to the targets. These sync operations (i.e.
3 In this scenario, if you are using different monitoring templates that contain different thresholds for the
metrics, after a database role change and templates are re-synced, you will not see a 'swap' of thresholds
between the primary and standby databases because they are separately associated with monitoring
templates that have different threshold values for the metrics.
template apply operations) are not done right away but are scheduled based on the synchronization
schedule. At the day/time specified by the synchronization schedule, all pending sync operations are
executed. Thus, you might want to think about specifying a synchronization schedule during off peak
hours (e.g. Saturday at 11 pm), when there is little or no impact to production operations. The
synchronization schedule allows you to specify both a start date/time as well as a frequency in days.
Figure 12. The Synchronization Schedule determines when scheduled sync operations (i.e. template apply operations) will be
performed by Enterprise Manager.
For example, to specify a schedule where sync operations occur only on Saturdays, specify a start date
that occurs on a Saturday and a frequency of 7 days. Because the synchronization schedule impacts all
sync operations, only an Enterprise Manager super administrator can specify or change the
synchronization schedule. Note that sync operations are scheduled only as needed, i.e. when the
conditions previously described occur. For more details on the synchronization schedule, refer to the
Implementing Administration Groups and Template Collections chapter of the Oracle Enterprise
Manager Cloud Control Administrator’s Guide 13c.
Define Roles for Different Job Responsibilities
Plan Job Responsibilities (Who can do what operations?)

In order to manage the group, you will need to think about the different job responsibilities as it
pertains to managing the group and its member targets. Here are some things to consider:
• Who can define group membership?
• Who can grant privileges on the group to other administrators?
• Who can do the following operations on the member targets:
o Define and apply monitoring settings
o Define and apply notification settings (e.g. who gets notified on events, etc.)
o Determine which events should have incidents created for them
o View and receive notifications for events/incidents
o Acknowledge and work on incidents
o Perform target blackouts for planned downtime activities
As mentioned earlier, there is a good amount of planning needed outside of Enterprise Manager to
capture your datacenter requirements and determine how these can be implemented in Enterprise
Manager. Planning discussions should also include defining job responsibilities.
For illustration purposes, here are some examples of some job responsibilities:
• Group Administrator
o Responsible for defining group membership and for granting privileges on groups to
other administrators
• Senior Administrator
o Responsible for adding and removing targets in Enterprise Manager, and for planning
and setting up monitoring settings for targets in Enterprise Manager. He is also
responsible for setting up rules related to creating incidents for events and sending
notifications based on the agreed upon plans.
• Target Owner
o For the targets he owns, he is responsible for setting monitoring settings, responding
to events/incidents, and for performing maintenance operations
• First Level Support
o Responsible for responding to events/incidents on targets
After you’ve defined the different job responsibilities, you will next need to understand the privileges in
Enterprise Manager required to support the various job responsibilities.
Understand Enterprise Manager Privileges to support job responsibilities

Enterprise Manager supports fine-grained privileges to enable more granular control over actions
performed in Enterprise Manager. The table below shows a (non-exhaustive) list of various job
responsibilities and the corresponding privilege in Enterprise Manager required to support these.
Table 1. Enterprise Manager Privileges to Support Job Responsibilities
JOB RESPONSIBILITY ENTERPRISE MANAGER PRIVILEGE
Create Administration Group Full Any Target

hierarchy Create Privilege Propagating Group
Edit Administration Group hierarchy • Full Any Target

• Create Privilege Propagating Group (if adding new target
property values as group criteria within a level of the
Administration Group hierarchy)
Delete Administration Group • Full Any Target

hierarchy
View entire Administration Group • View Any Target

hierarchy in Group Administration Note: Administrators who have privileges to only a subset of
pages the groups can view these groups in the Groups list page
accessible via Targets → Groups.
Use Monitoring Templates • Create Monitoring Template resource privilege is required

to create new monitoring templates.
• If you are using monitoring templates created by another
user, then you will need at least View privileges on those
monitoring templates.
• If you are using corrective actions, you will need the
Create Corrective Actons resource privilege to create
corrective actions or use any corrective action with your
monitoring template.
Use Template Collections • Create Template Collection (to create new Template
Collections)
• View Template Collection on specific Template Collection
to view/associate the Template Collection created by
another user
• View Any Template Collection to view/associate any
Template Collection
• Full Template Collection on specific Template Collection to
edit/delete the Template Collection created by another
user
Associate a Template Collection with • Manage Template Collection Operations on the group
an Administration Group (this includes Manage Target Compliance and Manage
Target Metrics privileges)
• View Template Collection on the Template Collection
Operations on the Administration

Group • Group Administration on the group
• Manage privileges on the group
(e.g. grant to other users)
• Configure Target (on the target to be added to the
• Add a target to an Administration Administration Group)
Group by setting its target
• Manage Template Collection Operations on the group
properties
• Perform a manual sync of the group
with the associated Template
Collection
Operations on the members of the

Administration Group
• Delete the target from Enterprise • Full on the target (Full also contains the privileges
Manager enumerated below)
• Operator on the target also contains all the privileges
enumerated below
• Set blackout for planned downtime o Blackout Target on the target
• Change monitoring settings o Manage Target Metrics on the target
• Change monitoring configuration o Configure Target on the target
• Manage events and incidents on o Manage Target Events on the target
the target o View on the target
• View target, receive notifications for
events or incidents
Create rules for managing events and • Create Enterprise Rule Set resource privilege
incidents on targets • Manage Target Events on the target on which the rule set
operates
For convenience, there is an out-of-box role called EM_TC_DESIGNER that contains the necessary
privileges required for creating Template Collections. Based on the job responsibilities you’ve defined,
you will then need to map the corresponding Enterprise Manager privileges needed to support the
various job responsibilities.
Use roles to manage user privileges

Privileges are ultimately granted to administrators to enable them to manage targets in Enterprise
Manager. While you can grant specific privileges to individual administrators, tracking and granting
privileges on many targets across many administrators easily becomes error-prone and an
administrative burden. Our recommendation is to define and use roles to manage the granting of
privileges to administrators. A role is a user-defined set of privileges typically containing the set of
privileges that you want to grant to a team of users. A role can contain other roles as well. For
example, you can create a First Line Support role containing the privileges needed for the
administrators to view and manage incidents on targets. Once this role is created, you can grant this
role to the appropriate administrators who will manage these incidents as part of their job
responsibility. If you need to change the set of privileges for your administrators, e.g. add new
privileges or remove privileges, then all you need to do is update the role. The updated set of
privileges in the role is automatically enabled for the administrators to whom the role has been granted.
Likewise, if new administrators are added, all you need to do is grant them the appropriate role(s)
instead of granting them individual privileges.
Using roles is one big step towards managing privileges. However, there is still the challenge of having
to keep the role updated with privileges on new targets as they are added to Enterprise Manager.
Privilege-propagating groups are meant to address this challenge and will be discussed next.
Leverage the privilege-propagating nature of Administration Groups
Administration Groups are privilege-propagating in nature. This means that a privilege on the
Administration Group that is granted to a user or a role automatically applies (i.e. “propagates”) to all
members of the group including any subgroups. If a new target is added to an Administration Group,
then because the Administration Group is privilege-propagating, the user or role that has privileges on
the Administration Group automatically gets privileges on the newly added target by virtue of it joining
the group. No additional work is needed for granting privileges on the new target. Thus granting target
privileges is much simpler because all you need to do is a one-time setup of granting privileges on the
group to a role.
In Figure 13 below, we have an Administration Group hierarchy. A role called “Senior Administrator
Role” has been granted Full on the PROD Administration Group. This means any user that has been
granted this role has Full privileges not only on the PROD group itself but all subgroups and all the
members of the subgroups Sales, Finance, Others under the PROD group. Similarly, a (lesser) privilege
Manage Target Events on the PROD group has been granted to the “First Line Support” role. This
means any administrator that has been granted this role has Manage Target Events on PROD group
and all its subgroups. In the future, if a new target is added to the Sales, Finance, Others subgroups
under PROD group, the user who has been granted “Senior Administrator Role” will automatically get
Full privileges on the newly added target and the user who has been granted “First Line Support” role
will automatically get Manage Target Events on the newly added target. Thus, roles and privilege
propagating groups provide benefits in economies of scale when it comes to managing a large number
of target privileges for a large number of users.
Figure 13. This shows the privilege-propagating nature of Administration Groups.
Create Roles for different job responsibilities

After you’ve planned the various job responsibilities and mapped these to the corresponding privileges
in Enterprise Manager, the next step is to create roles in Enterprise Manager containing privileges
required for each job responsibility. In our example below, here are the various roles that need to be
created for each job responsibility. Note that when it comes to privileges on targets in the
Administration Group, the recommendation is to grant the privilege on the Administration Group and
not on individual targets in order to leverage the privilege propagating nature of Administration
Groups:
Table 2. EXAMPLES OF ROLES YOU CAN CREATE FOR DIFFERENT JOB RESPONSIBILITES*
JOB RESPONSIBILITY ROLE IN ENTERPRISE *PRIVILEGES IN THE ROLE

MANAGER (MINIMUM SET)
Group Administrator GROUP_ADMIN_ROLE Group Administration on the group

Responsible for defining
group membership and
for granting privileges on
the group to other
administrators
Senior Administrator SENIOR_ADMIN_ROLE • Add Any Target

Responsible for adding • Create Enterprise Rule Set
and removing targets in • Operator on the group
Enterprise Manager, and
• Create on Job System
for planning and setting
up monitoring settings for • EM_TC_DESIGNER role
targets. He is also
responsible for setting up
rules related to creating
incidents for events and
sending notifications.
Target Owner TARGET_OWNER_ROLE • Operator on the Administration

For the targets he owns, Group(s) that he is managing
he is responsible for • Create on Job System
setting monitoring settings, • View Any Monitoring Template
responding to
View on the Template Collection(s)
events/incidents, and for
associated with the group(s) he is
performing maintenance
managing
operations
First Level Support FIRST_LEVEL_SUPPORT • Manage Target Events on the

Responsible for appropriate Administration
responding to Group(s)
events/incidents on • Blackout Target on the appropriate
targets. As part of Administration Group(s)
operational procedures, he
is allowed to blackout a
target that is down.
* The privileges listed in Table 2 represent the minimum set of privileges in the role. Additional
privileges can be added based on other responsibilities. Also note that you will need to have Super
administrator privileges to create roles.
To create roles, log on as an Enterprise Manager super administrator. Go to Setup → Security → Roles.
Select the Create button and follow the steps in the Create Role wizard to create the role with the
necessary privileges.
There may be cases wherein the Administration Groups do not exactly fit the group requirements as far
as managing privileges are concerned. For example, for an Administration Group, you might want to
grant privileges for only a subset of the targets in the group. For recommendations on how to resolve
this, refer to the section called Using Administration Groups for Other Group Operations in the latter
part of this paper.
Assign Roles to your Enterprise Manager administrators

Once roles have been defined, you can now grant these roles to your Enterprise Manager
administrators. This can be done in several ways:
• When creating/editing an Enterprise Manager administrator, you can assign role(s) as part of the
“Create/Edit Administrator” wizard.
• As part of creating/editing a role, the “Create/Edit Role” wizard allows you to choose administrators
to whom you would like to grant the role.
• When creating/editing administrators using the Enterprise Manager Command Line tool (EM CLI)
create_user or modify_user, you can specify the roles granted to the user. Also EM CLI verb
grant_roles also grants a role to a user (administrator).
Set up Rules to Manage Events and Incidents

The last phase in setting up monitoring involves setting up rules in Enterprise Manager to automate the
operational processes for managing events based on your requirements. For example, one
requirement might be to send page notifications for critical events and send email notifications for
warning events. Or another might be to open helpdesk tickets for target down events. Rule Sets
enable the automation of these notification actions as well as other actions on events and incidents.
This phase involves following:
• Understanding rule sets and rules
• Using groups in rule sets
• Auto-creating incidents using rules
• Planning your rules to leverage Enterprise Manager rule features
• Implementing the rules in Enterprise Manager
Each of these will be discussed in further detail. But before we get into these details, it is first important
to get a basic understanding of events and incidents in Enterprise Manager.
Events and Incidents

An event is a significant occurrence, typically on a managed target, that has been detected and raised
by Enterprise Manager. For example, a target that is down causes a ‘target down’ event to be raised.
When disk usage is nearing its capacity, a ‘disk full’ event is raised. If a job fails, then a ‘job failed’ event
is raised. There are different types of events in Enterprise Manager4 and sometimes you might see
multiple events raised that pertain to the same underlying issue. For example, if a host target is under
heavy load, then you could potentially see a CPU Utilization(%) event, Memory Utilization(%) event and
Swap Utilization(%) event on the host all trigger within minutes of each other. Administrators can get
easily overwhelmed by the volume of events raised. Thus, it is desirable to be able to narrow down
these set of events into the subset of actionable events that need to be addressed (because they impact
your business applications) and also identify which of these pertain to the same issue and thus can be
logically managed collectively as one unit. It is also desirable to be able to assign, prioritize and track
the resolution of these important events. In Enterprise Manager 13c, we enable this through Incident
Management features. The primarily goal of incident management is to be able to monitor and resolve
disruptions to services as quickly and efficiently as possible. From the set of events raised, one should
create incidents for significant, actionable events (e.g. target down event) or for a combination of
significant related events that all pertain to the same issue (e.g. one incident for CPU Utilization(%),
Memory Utilization(%), Swap Utilization(%) events on the same host). An incident is thus an object in
Enterprise Manager containing either a significant actionable event or a combination of actionable
related events that pertain to the same issue.
Figure 14. This is an example of an incident containing one event.
In Figure 14 above, the incident contains one target down event. In Figure 15 below, the incident
contains multiple events. You can also combine multiple events into the same incident using the
Incident Manager console or using event rules. The latter method will be discussed in the section
Minimizing Event Storms by Compressing Multiple Events into a Single Incident. The severity of the
incident is the worst-case severity of the events it contains.
Figure 15. This is an example of an incident containing two events.
4 There are multiple event types supported. These include: Metric Alert, Target Availability, Job Status Change, Compliance Standard
Score Violation, Service Level Agreement Alert, Metric Evaluation Error, JVM Diagnostics Threshold Violation, etc. For more details
refer to the Incident Management chapter of the Oracle Enterprise Manager Cloud Control Administrator’s Guide 13c.
Once an incident is created, a rich set of incident lifecycle management features are available in
Enterprise Manager to help manage the resolution of incidents. These include the sending of
notifications, assigning ownership, acknowledging an incident, tracking its resolution status,
prioritizing, escalating long running incidents, adding annotations, etc. Thus incidents (and not
individual discrete events) are the primary way to monitor and manage events raised in Enterprise
Manager. For more details on incidents refer to the “Using Incident Management” chapter of the
Oracle Enterprise Manager Cloud Control Administrator’s Guide 13c. Additional recommendations on
managing incidents will be discussed in subsequent sections of this paper.
Understand Rule Sets and Rules

Automating actions on events and incidents in support of operational processes is an important part of
any scalable monitoring solution. In Enterprise Manager, this is supported using rule sets and rules. It
is important to get a good understanding of rule sets and rules before you can leverage them in your
Enterprise Manager deployment. A rule is an instruction for Enterprise Manager to take action on an
event or incident or problem. (A Problem is another type of object used in Enterprise Manager to
manage critical errors in Oracle software. These will be discussed separately). Actions include sending
of email notifications, opening of helpdesk tickets, creation of incidents for events, assignment of
incidents, etc. A rule set is a set of rules that operate on a common object such as a group of targets.
A rule consists of:
• Criteria: specified set of events or incidents on which the rule will operate. Examples are: all target
down events on all targets in the group, specific metric alert (events) such as host Filesystem Space
Available(%), database Tablespace Space Used (%), etc.
• Actions: one or more actions that Enterprise Manager should take on the specified set of
events/incidents in the criteria. Actions in turn can be ‘conditional’, i.e. only execute the action if a
specified condition is met. For example, your rule criteria could be “all incidents of fatal severity” and
the action could be: if “the incident has been opened for more than 48 hours” (condition part), then
set the escalation level to 2 (action part).
If you have several rules that apply to the same object (such as a group), then they should be combined
in a rule set. Rules within rule sets are executed in a certain order. By default, they will be executed in
the order in which they are defined, but they can be re-ordered as needed. Rule sets are also executed
in a specified order. By default, they will be executed in the order in which they are defined, but they
can also be re-ordered. Creation and use of rule sets are defined in detail in “Using Incident
Management” chapter of the Oracle Enterprise Manager Cloud Control Administrator’s Guide 13c. It is
recommended that you refer to the documentation for details on rule set features. This paper will focus
on how you can set up rule sets in support of your operational processes.
Use Groups in Rule Sets

For monitoring purposes, the best practice recommendation is to specify a group (or groups) of targets
as the object of the rule set. In a hierarchy of groups, i.e. group containing other groups, specify the
highest-level group for which the rules apply. Each rule in the rule set operates on the applicable
member targets of the group including members of any subgroups. It is also recommended that any
group you use for rule sets be included in only one rule set and not in multiple rule sets. Putting all the
rules related to the same group into one rule set makes it much easier to track and manage all actions
on the group because it’s centrally defined in one place. The group you use in your rule set should
contain targets that have common requirements for notification as well as common requirements for
actions on events and incidents (such as assignment of incidents, etc.). As the group later expands and
additional targets are added, the rule set will automatically apply to the newly added members without
further modification of the rule set. If an Administration Group defined in the prior sections can serve
the purpose of being used for rule sets (i.e. all members of the Administration Group have common
requirements for notifications and event/incident management actions), then the Administration
Group should be used as the target of the rule set.
Auto-create Incidents Using Rules

Since incidents are the recommended way of monitoring and managing events, it is recommended that
rules be used to automatically create incidents for events that are important to be managed. In a rule
set, the rules that create incidents should typically be the first set of rules in the rule set. There is an
out-of-box rule set called Incident management rule set for all targets that automatically creates
incidents for a subset of important events (such as target down availability events). However, it most
likely needs to be adjusted based on your particular Enterprise Manager deployment needs. It is
recommended that you review this out-of-box rule set to see if it meets your requirements. If it does
not, you can do a ‘create like’ on the rule set and change the new rule set as needed. If you do so,
remember to disable the original out-of-box rule set.
Minimizing Event Storms by Compressing Multiple Events into a Single Incident

When using rules to create incidents for events, you should consider using the event compression
feature to compress or combine multiple related events into a single incident. Here are some examples
of where you may use this feature:
• Compress or combine target down events for all RAC instances that belong to the same RAC into
one incident
• Compress or combine configuration standard violation events into one incident
• Compress or combine all metric collection error events for a target into one incident
The overall severity of the incident is the worst severity across all of the events it contains. For
example, if the incident contains an event of critical severity and an event of warning severity, the
severity of the incident is ‘critical’. If you need to send notifications for these events, send notifications
for the incident, and not for the individual events. Using the compression feature in this way reduces
possible ‘event storms’ when these events happen at about the same time, and enables you to work on
these events in a more manageable way.
Plan Your Rules

Rules are meant to automate the operational processes for managing events and incidents. Before
creating the rules in Enterprise Manager, it is important to take time to plan the structure of your rules.
For purposes of illustration, it is assumed that the set of rules that administrators would like to set up
follow this type of pattern:
Rule requirements for the production group PROD:
1. For target down events, these should be treated as highest priority. Page the administrators.
(Note: Target down events are of ‘fatal severity’.)
2. For specific set of metric alert (events) on the databases in the group, send a notification based
on the severity: if it is critical severity, page the administrators; if it is warning severity, email
the administrators.
3. For specific set of metric alert (events) on hosts in the group, send a notification based on the
severity: if it is critical severity, page the administrators; if it is warning severity, email the
administrators.
4. For specific set of metric alert (events) on WebLogic Servers in the group, send a notification
based on the severity: if it is critical severity, page the administrators; if it is warning severity,
email the administrators.
5. If any job against the group fails, page the administrators. (Note: Job failure events are of
‘critical’ severity).
6. If the target down event isn’t resolved within 24 hours, escalate to level 1 and email the
manager.
7. If the target down event isn’t resolved within 72 hours, escalate to level 2 and email the
manager.
When transforming the above rule requirements into actual rules in Enterprise Manager, here are some
guidelines you can use:
1. Put all the rules that pertain to the common target (e.g. group) into one rule set instead of
across multiple rule sets. It will make it easier to manage the rules that pertain to the same
target in one place and avoid potential duplication of actions across rule sets.
2. Incidents should typically be created for the events specified in the rule. Since there are
operational requirements to send notifications for these events, then this means these events
are actively managed and thus incidents should be created for these. Also think about which
events can be logically compressed (or combined) into one incident. Rules that create
incidents for events should typically be the first set of rules in the rule set.
3. Once incidents are created, subsequent actions in the rule set such as notifications, etc. should
typically be performed on the incident instead of on the event. One exception would be in the
case where interested parties are interested in receiving notifications for specific events. For
example, business owners might want to be notified of ‘target down events’ for targets that
impact their business applications. For these cases where the primary interest is to be notified
if such events occur, you can create rules on the specific events of interest and specify
notification actions for the interested parties. Extending our example, if business owners
wanted to be notified if any of their important targets went down, you could create a rule on
the ‘Target Availability’ down event and the action of the rule would be to send email
notification to the business owners.
4. If there are common actions across rules (e.g. for rules #1 through rule #5 in the rule
requirements list above, actions are to send notifications), consider creating a separate rule for
these actions and have this new rule be executed after the other rules. If it is set up this way,
then should there be any future changes to the actions, you will only need to make the
changes in one rule instead of multiple times across several rules. In our example, the actions
that are common across rules #1 through rule #5 are to send page notifications if the severity
is fatal or critical, and send email notifications if the severity is warning. Instead of specifying
the notification actions for each of the five rules, you can create a separate rule for the
notification actions. If there is any change needed in the notification action (such as a change
in the recipient), then you will only need to make the change in this separate rule.
5. Consider combining rules that are similar in nature and only differ by different parameters
(e.g. rules #6 and #7 above) by leveraging conditional action support in rule actions. In our
example, rules #6 and #7 both apply to target down events. They differ in the length of time
the event has been opened and subsequent actions based on this. This can be combined into
one rule with two conditional actions: the rule will apply to target down events (or incidents
containing target down events) and the conditions for the actions will be based on ‘duration’
i.e. how long the incident has been opened. The first action will have a condition based on the
incident being open for at least 24 hours, and the second action will have a condition based on
the incident being open for at least 72 hours.
Applying the above guidelines to the rule requirements, we now have these rules:
Rules for the production group PROD:
1. Create incidents for the following events: target down events, specific metric alert (events) for
database, host, WebLogic targets, and job failure events. For target down events from RAC
instances, compress these into a single incident, one incident per RAC target. Also set the
priority of these incidents to Urgent.
2. Send notifications for incidents as follows: send page notifications to administrators for
incidents of fatal severity or critical severity; send email notifications to administrators for
incidents of warning severity. (Note: target down events are the (only) events with fatal
severity; backup job failure events have critical severity). As a variation of this rule, you could
have different recipients for the notifications based on target type, i.e. the DBA team gets
notified for incidents on database targets, the middleware team gets notified on incidents on
WebLogic targets and the system administrators get notified for incidents on host targets.
3. For incidents of fatal severity (i.e. incidents for target down events):
a. If the incident is still open after 24 hours, then escalate incident to level 1 and email
the manager.
b. If the incident is still open after 72 hours, then escalate the incident to level 2 and
email the manager.
Which type of rule should I use?

Now that you have planned the rules that you would like to setup, when you define these rules in
Enterprise Manager, there is a choice of the type of object on which the rule will operate: event or
incident, or problem. General recommendations on which object to use for rules are described in
Table 3.
Table 3. RECOMMENDATIONS FOR USES OF RULES
TYPE OF RULE PURPOSE
Rule on Events • Create incidents based on one or more events

• Create helpdesk tickets for incidents (create incident based on
the event, then create the ticket based on the incident)
• Send events to third party management systems
• Send email for events you’re interested in (e.g. sending email to
users such as business owners for specific events of interest)
Rules on Incidents • Automate incident workflow operations (e.g. assignment of

incident, prioritization, etc.)
• Send notifications on incidents
• Create helpdesk tickets for incidents. You can optionally specify
conditions for when the ticket should be created (e.g. create ticket
if incident is escalated to level 2)
Rules on Problems • Automate problem workflow operations (e.g. assignment,

prioritization, etc.)
• Send notifications on problems
The rule set for the production group PROD should thus be defined as follows in Enterprise Manager:
Rule Set name: Rule set for production group PROD

Target: PROD Group
Rules: (See Table 4 below)
Note the targets and events defined in the rule criteria in Table 4 below are scoped to the PROD group
which is the target specified in this rule set.
Table 4. RULES IN THE RULE SET FOR PRODUCTION GROUP PROD
EXECUTION RULE APPLIES RULE CRITERIA RULE ACTION(S)

ORDER TO…
1 Events (All target down events from all • Create incident, compress
database instances) events into an incident by
• Event type = Target ‘Target’ with same ancestor
Availability AND type ‘Cluster Database’
• Event = (Target Type = • Set incident priority = Urgent

Database Instance, Availability Note: This compresses all target
State = ‘Down’), (Target Type down events from RAC Instances
= Cluster Database, into an incident, one incident per
Availability State = ‘Down’) RAC. If the RAC target is also
down, its event is also included in
the same incident.
ORDER TO…
Target down events from Single

Instance databases (that have the
same target type as RAC
instance) will have individual
incidents created for them.
2 Events (All target down events across Create incident,

all WebLogic servers) Set incident priority = Urgent
• Event type = Target
Availability AND
• Event = ‘Target down’ for all
WebLogic servers
3 Events (Specified set of metric alerts for Create incident

databases, hosts, WebLogic
servers)
• Event type = Metric Alert AND
• Metrics = <specified set of
metrics> with warning or
critical severity
4 Events (All job failure events) Create incident

• Event type = Job Status
Change AND
• Job Status = Problems
5 Incidents (All incidents of fatal, critical or • If severity is critical, page

warning severity) <specified set of system
• Severity in (Fatal, Critical, administrators>
Warning) AND • If severity is warning, email
• Target Type = Host <specified set of system
administrators>
6 Incidents (All incidents of fatal, critical or • If severity is fatal or critical,

warning severity) page <specified set of DBAs>
• Severity in (Fatal, Critical, • If severity is warning, email
Warning) AND <specified set of DBAs>
• Target Type = Database
7 Incidents (All
incidents of fatal, critical or • If severity is fatal or critical,
warning severity) page <specified set of
• Severity in (Fatal, Critical, middleware administrators>
Warning) AND
ORDER TO…
• Target Type = WebLogic • If severity is warning, email

Server <specified set of middleware
administrators>
8 Incidents (All incidents of fatal severity) • If incident is open for at least 24

Severity in (Fatal) hours, set incident escalation
level to 1 and email <specified
manager>.
• If incident is open for at least 72
hours, set incident escalation
level to 2 and email <specified
manager>.
Other Considerations for Planning Rules

Rules were designed to offer much flexibility in specifying the conditions and actions that the rule
should take. As such, there could be possibly more than one way to define rules that meet your
operational requirements. Here are additional options you might want to consider when defining rules:
1. Which rule should create the incident?
An event can be part of at most one incident. Hence if there are multiple rules that create an incident
for the same event, then the rule that is executed first will create the incident containing the event.
Subsequent rules that contain actions to create an incident for the same event will not be executed
because an incident has already been created for the event. However, other applicable actions (such as
sending notifications) will still be executed. Hence to avoid surprises, it is important to plan which rule
set will contain rules for creating incidents.
2. Execution order matters if rules operate on the same set of events and incidents
Incident workflow attributes such as owner, priority, etc. can be changed via rules. Thus, it is technically
possible for a set of rules in the same rule set to set the same incident attribute to different values. If
this happens, the last one wins, i.e. the final value of the incident attribute is based on the last rule that
was executed. You might have requirements that require the setting of incident attributes to different
values based on certain conditions. In this case, think about putting rules that apply to broader criteria
first, and rules that apply to more specific criteria afterwards. For example, if your requirements are to
assign incidents on database targets to the DBA-TEAM user and all other incidents to the SUPPORT-
TEAM, then you can set the rules up this way:
Rule1: For all incidents, set owner to SUPPORT-TEAM
Note: The rule above is the rule with broader criteria.
Rule2: For all incidents on database targets, set owner to DBA-TEAM
Note: The rule above is the rule with more specific criteria.
Using the above rules, if the incident is on a database target, then even if the owner was initially set to
SUPPORT-TEAM (per Rule 1), it will be later changed to DBA-TEAM per Rule2 that was executed after
Rule1.
Finally, if you have rules containing duration-based criteria in the actions part of the rule (e.g. if incident
is of Urgent Priority and has been open for at least 24 hours…), then try to put these rules as last in
execution order. It is likely that the conditions specified in these rules are based on incident attributes
(e.g. priority, etc.) and since such attributes can be changed by other rules, putting duration-based rules
as last ensures the values of the attributes that they rely on are the latest values.
3. Broader vs. more specific rule criteria
In the rule set defined in Table 4 above, the rule to generate incidents for metric alerts involves
specifying individual metrics. One advantage to this approach is that you will know exactly which
metric alert (events) generate incidents for your group. However, if you want to monitor additional
metrics for any of the targets in the group, and you would like to create incidents for these, you will
need to remember to update the rule by adding the additional metrics. An alternative solution to this is
to not specify individual metrics in the rule criteria but to expand the rule criteria to all metric alert
events of warning or critical severity as shown in Figure 16 below:
Figure 16. This rule applies to metric alerts (events) that are in warning or critical severity.
Again, note that this will apply to all warning and critical metric alerts for the PROD group and not in the
entire Enterprise Manager deployment. The advantage to this approach is that it will automatically
cover any additional metric alert (events) on the group, i.e., there is no need to update this rule should
you decide to monitor any additional metrics for targets in the group. However, this assumes that
there is a fair amount of control over setting the appropriate thresholds of the metrics across all targets
in the group (hopefully done through the Administration Group – Template Collection feature).
Otherwise if you have some spurious metric alert (events) due to incorrect metric thresholds, this rule
will end up creating some unnecessary (unwanted) incidents for your group.
4. Email notifications: the ‘To’ and ‘Cc’ lines
When specifying the recipients of email notifications, you will notice there are 2 fields on which you can
specify the recipients: “E-mail To’ and ‘E-mail Cc’
Figure 17. This shows the options for the e-mail recipients in a rule.
The ‘E-mail To’ line should be used to specify the primary recipients of the email notifications, and
these are typically the administrators who are expected to take some action in response to the
event/incident/problem for which they are being notified. The selection of these recipients are
determined as part of the planning process and not an ad hoc choice for any individual administrator.
As such, changing the recipient in the ‘To’ line can only be done by changing the rule itself (which is
restricted to the rule creator and other designated administrators as will be discussed in the next
section). The ‘E-mail CC’ line should be used for other interested parties (e.g. business owners, etc.)
who would like to be notified of the event/incident/problem’s occurrence. As such, these users can
subscribe or un-subscribe to these rules based on their own interest.
5. Email recipients need privileges on the targets
When creating rules that send email/page notifications to users, these users will only be able to receive
notifications if they have privileges (i.e. at least View privileges) on the targets on which the
events/incident have been created.
6. Determine who creates the rules in Enterprise Manager
When actions on events/incidents (and problems) are specified in rules, these are executed by
Enterprise Manager using the privileges of the Enterprise Manager administrator who created the rule.
Thus, it is important to ensure that the creator of the rule has the necessary privileges on target(s) of
the rule set for the actions to be executed successfully; otherwise, the rule action will fail. If the target of
the rule set is a privilege propagating group, then the privilege can be granted once on the group and it
will automatically be applied to all members of the group. For example, if one of the rule actions is to
‘Create an incident’ for an event, then the rule creator must have at least Manage Target Events
privilege on the target on which the event has occurred. The checking of privileges of the rule creator is
done at rule execution time so it is important that you proactively check that the rule creator has
sufficient privileges (e.g. Manage Target Events) on the targets in the rule, otherwise rule actions will
fail. For a complete list of rule actions that use the privileges of the rule creator, refer to the “Incident
Management” chapter of the Oracle Enterprise Manager Cloud Control Administrator’s Guide 13c.
7. Audit changes to the rules
Since rules are an integral part of your monitoring and incident management processes, consider
enabling auditing of your rules to help you can keep track of who and when rule changes were made.
You can use EM CLI update_audit_settings to enable such operations to be audited. Refer to the
section on “Configuring and Managing Audit” of the Oracle Enterprise Manager Cloud Control Security
Guide for details on enabling audit for incident rule set changes.
8. Verify rule setup using ‘Simulate Rules’ feature
You can use the ‘Simulate Rules’ feature to help you verify your rule set setup. Simulate Rules can be
accessed in the main Incident Rules page. The simulate rule feature will list out all the actions that
Enterprise Manager would take if a specified event occurs. To use this feature, you will first be asked to
choose an event from a list of existing events. Then you can run the simulation. The output of the
simulation is a list of rules and actions that the rules would take if the event occurs. The actions will not
actually be executed, thus providing a safe way for you to test your rules. If there are missing rules or
actions in the simulation output, then that indicates some issue in the rule setup. It could be due to
missing privileges on the rule creator (e.g. Manage Target Events privilege is required to create an
incident) or perhaps the event conditions in the rules are incomplete.
Implement the Rules in Enterprise Manager

Once the rules have been planned, they can be implemented, i.e. entered in the Enterprise Manager
console (Setup → Incidents → Incident Rules). Note that you will need the Create Enterprise Rule Set
resource privilege to create these rules. All the rule sets that have been discussed thus far are of type
“Enterprise”. Enterprise rule sets are used to implement your IT operational processes (send email,
open helpdesk tickets, escalate incidents, etc.). Because these rule sets can perform a wide variety of
actions, the creation of these rule sets is protected by a resource privilege called Create Enterprise Rule
Set. It is recommended that specified persons are designated to create these rules on behalf of the
team (this is done as part of the planning of job responsibilities discussed in prior sections of the
paper). If multiple people are designated as having this responsibility, then once one person has
created the rule set, the other users can be made Co-authors of the rule set. This entitles them to also
edit the rule set as needed. Finally, it is also possible to enter the rule sets in advance of actual
operations, i.e. in advance of having the rules be operational in your environment. This can be done by
disabling the rule set and re-enabling it once you are ready to go live.
Benefitting From Economies of Scale: A Fully Automated and

Scalable Monitoring Setup
At this point you’ve setup the groundwork for an automated and scalable monitoring solution. The
‘cost’ or effort of setting up new target monitoring is now very minimal. Whether you’re adding one
new target or hundreds of new targets to Enterprise Manager, all you need to do as far as monitoring
setup is concerned is one step: set the target properties of the target such that it matches the
membership criteria of the Administration Group to which is should belong. Once you’ve done this,
everything else is automatically handled by Enterprise Manager:
• The target is automatically monitored using the appropriate monitoring settings.
o This happens because the target is automatically added to the appropriate Administration
Group. Once it is in the Administration Group, the monitoring template from the
associated Template Collection is automatically applied to the target.
• Administrators can start managing the target because they have the appropriate privileges to
manage the newly added target.
o This happens because the role that has been granted to them contains privileges on the
Administration Group and since Administration Groups are privilege-propagating, it
automatically includes privileges on the newly added target.
• Rules for creating incidents, sending notifications, and other operational procedures automatically
include the newly added target.
o This happens because the Administration Group has been specified as the target for the
rule set and newly added targets for the group are automatically included in the rule set.
Figure 18. Monitoring set up for new targets is fully automated.
MANAGING INCIDENTS
Once monitoring has been setup, events will be raised and incidents created for these events. Incident
Manager (available within the Enterprise Manager console) provides a centralized way to manage all
these different incidents. Here are some guidelines on how you can effectively manage incidents:
1. Leverage views in Incident Manager
The views in Incident Manager enable you to focus on the subset of incidents you’re interested in.
There are several out-of-box views that cover a common set of incidents that a datacenter might be
interested in. For example, an administrator can use the out-of-box view ‘My open incidents and
problems’ to view all the incidents and problems assigned to him. This list can be sorted by priority so
that he can work on the most important ones first. You can also create additional (custom) views to
filter the list of incidents based on your specified search criteria. A good example would be to create a
view containing all incidents for an important group of targets (e.g. production group PROD). These
custom views can also be shared with other users. Hence you can designate a user in your team to
create views on behalf of the team, then share the custom views created.
2. Automate incident assignment
Automate the assignment of incidents to administrators using rules if the set of incidents that should
be assigned to administrators is well-defined (e.g. all host incidents are assigned to user A, all database
tablespace incidents go to user B, etc.)
3. Manage by priority
The priority field of an incident should be leveraged in order to provide guidance on which incidents are
more important and thus should be managed first. If the semantics for determining which incidents are
the highest priorities can be determined in advance, you can use rules to automatically set the incident
priority. In Incident Manager, when administrators are working from a list of incidents (within a view), it
is recommended that they sort the incidents by priority to manage the more important ones first.
4. Working off a queue of incidents
If you operate by having a team work off a common queue of incidents, you can emulate the ‘queue of
incidents’ for the team by creating an Enterprise Manager user to represent the team. For example, you
can create a user called ‘DBA-TEAM’ to represent the DBA team. If there is a well-defined set of
incidents that need to be owned by the DBA team, then create a rule to auto-assign these incidents to
user DBA-TEAM. This can be done as part of the rule that creates the incident. (For example, create a
rule on events and choose specific database metric alerts; in the actions part of the rule, create an
incident and assign the incident to DBA-TEAM.) In Incident Manager, create a view to filter the list of
incidents to all incidents owned by user DBA-TEAM. The DBA team can then use this view to look at all
the incidents assigned to their team. When they actually want to work on an incident, they should take
ownership of the incident so that the owner field is set to their own Enterprise Manager user account.
If there is an additional requirement to send email notifications to the DBA team when an incident is
assigned to them, then you can specify the DBA-TEAM as the recipient of the email notification in your
event/incident rule and specify an email distribution list as the email preference for the DBA-TEAM
user. The email distribution list should include the email addresses of the members of the DBA team.
5. Automate incident escalation
Automate the escalation process using rules (e.g. set escalation level and send notifications to
appropriate people based on type of incident and duration). In the rule, you can specify a duration
condition on the incident that determines when the incident should be escalated.
6. Check for missing incidents
Occasionally review the set of events in the view Events without incidents. There may be some
important events for which no incidents have been created and thus are left unattended. If you find
such events, then you can manually create an incident for them in Incident Manager. Also consider
creating a rule for these to automatically create an incident for the event the next time it occurs.
7. Leverage incident resolution status to track the progress of your incidents
The ‘status’ field of an incident (or problem) allows you to track the resolution of the incident. All newly
created incidents are set to ‘New’. As administrators work on the incident, they should change the
status appropriately (e.g. Work In Progress). This enables you to see how much (or how little) progress
has been made to the resolution of an incident. Additional values for the ‘status’ can also be added to
support your operational processes. For example, you can add values such as ‘Waiting on Vendor’ or
‘Waiting on Subject Matter Expert’ etc. To add additional values, use the EM CLI verb
create_resolution_state. You will need to have Enterprise Manager super administrator privileges to
execute this verb.
Many incidents in Enterprise Manager cannot be manually cleared. They can only be cleared if their
underlying event has been cleared by the agent. If an administrator believes he has fixed the incident
but is waiting for Enterprise Manager to clear the event (and incident), he can set the incident status to
‘Resolved’. Once the event (and incident) is cleared by the agent, the status will be automatically set to
‘Closed’.
8. Use Lifecycle Status target property to identify the most important targets
The Lifecycle Status target property is meant to designate the operational status and importance of a
target. Possible values are (in order of descending priority): Mission Critical, Production, Staging, Test,
and Development. If your Enterprise Manager site is under heavy load, then events from the higher
priority targets (e.g. Mission Critical and Production targets) are processed ahead of the lower priority
targets (e.g. Development targets). Target Availability events are always processed ahead of other
types of events. Set the appropriate value for the Lifecycle Status target property to your important
targets to ensure their events are processed ahead of others in case Enterprise Manager is under heavy
load. You can set the value of target properties using the Target Properties page which is accessible
from the target’s menu (Target menu → Target Setup → Properties). Or, to do this operation in bulk,
use EM CLI set_target_property_value. You cannot add or remove values from pre-defined set of
values for Lifecycle Status. However, you can change the display name using the EM CLI verb
modify_lifecycle_stage_name.
ADDITIONAL MONITORING REQUIREMENTS AND

RECOMMENDATIONS
Here are some additional common requirements, issues and recommendations related to monitoring:
Auto-fixing Alerts using Corrective Actions

There are certain types of alerts that have standard remediation actions. For example, if a listener is
down, the remediation action would be to restart the listener. It is more efficient to automate these
remediation actions in Enterprise Manager rather than to have these manually executed by on-call
staff. To automate these actions, you can use Corrective Actions. Corrective Actions are jobs that are
executed by the agent in response to an alert. Enterprise Manager provides some built-in corrective
actions such as: Startup Listener, Startup Database, Add Space to Tablespace. You can also specify
custom logic using the OS Command or SQL Script corrective actions.
Corrective actions are specified as part of a target’s metric threshold settings. When corrective actions
are executed, it must run using the credentials of a specified user. It is recommended that corrective
actions are included as part of the monitoring template that is in turn part of the Template Collection
that is associated with the Administration Group. If it is setup in this recommended way, then when the
corrective action executes, it will use the preferred credentials of the Enterprise Manager user who
associated the Template Collection with the Administration Group.
For more details on using corrective actions, refer to the “Utilizing the Job System and Corrective
Actions” chapter of the Oracle Enterprise Manager Cloud Control Administrator’s Guide 13c.
Monitoring When Enterprise Manager is Under Planned
Maintenance
Occasionally, administrators need to bring down Enterprise Manager’s Oracle Management Service and
Repository for planned maintenance. While the duration of this maintenance window may be short,
incident management and notification features are not available during this downtime and yet
businesses may require 24x7 continuous monitoring of their mission critical targets. In order to meet
the continuous monitoring requirements, you can use Always-On Monitoring. Always-On Monitoring is
a separate, complimentary application that can be configured to receive alerts from agents and send
email notifications for alerts. In Enterprise Manager 13.4, it will send email about the most important
alerts: Target Availability and Metric Alerts. You can configure it include additional alerts (e.g. metric
errors) if needed. For more details, refer to the “Always-On Monitoring” chapter of the Oracle
Enterprise Manager Cloud Control Administrator’s Guide 13c.
Sharing the Administration Group Hierarchy across Different

Teams
In some datacenters, one Enterprise Manager deployment could be shared across different teams that
are responsible for managing their own set of targets. One way for different teams to share the same
Administration Group hierarchy is to have the first level of the hierarchy be used to divide the targets
based on the teams that manage them. As a simple example, say the DBA team is responsible for the
databases and listeners, the Middleware team is responsible for the WLS (WebLogic) targets, and the
Sysadmin team is responsible for the host targets. The first level of the Administration Group
hierarchy can be based on the Target Type property such that the target types owned by a team are
part of the same group:
Figure 19. The same Administration Group hierarchy can be shared by different teams.
Additional levels, if needed, can be added to the Administration Group, but they would have to be the
same across all groups. For example, another level based on Lifecycle Status can be added to further
divide the targets for each team into production and non-production targets if there are different
monitoring settings for production and non-production targets. Since the levels must be the same
across all groups, these levels would need to be discussed and agreed-upon by the various teams.
However, it should be noted that all the levels need not be used by all teams. For example, in the
Administration Group hierarchy below (Figure 20), a Lifecycle Status level has been added to create
Production and Non-Production groups for each team:
Figure 20. This Administration Group shared across different teams has the same levels across all groups.
This enables each team to define different monitoring settings for Production targets and Non-
Production targets. In this example, it is important to note two points:
• If a team, say the Sysadmin Team, has the same monitoring setting for all host targets, then they do
not have to use the additional Production and Non-Production groups for template application
purposes. They can create a Template Collection containing the monitoring template for host targets
and associate the Template Collection with the group ‘Sysadmin Team Targets’. This template will
then apply to both subgroups Production and Non-Production.
• In order for a target to join the Administration Group hierarchy, it must match all the criteria used in
defining the hierarchy. In an Administration Group hierarchy, non-group targets are always found in
the lowest level of the hierarchy. So, in our example hierarchy, the target must match both the target
type and lifecycle status criteria in order to join the group. For host targets managed by the Sysadmin
team, even if there is no difference in monitoring settings between production and non-production
targets, the targets’ lifecycle status must be set to match the lifecycle criteria of the Administration
Group for the host target to join the Administration Group.
Finally, another benefit of using the first level hierarchy to divide targets by teams is ease of use of
managing privileges. If all members of a team have the same level of privileges for all targets they
manage and only view privileges (or no privileges) on the other targets that they don’t manage, then a
role can be created that grants the appropriate level of privilege on the first level groups. For example,
the role created for the DBA Team can consist of Full privileges on the ‘DBA Team Targets’ group, View
privileges on the ‘Sysadmin Team Targets’ group, and perhaps no privileges on the ‘MW Team Targets’
group. This is depicted in Figure 21 on the next page.
Figure 21. A role containing privileges on an Administration Group automatically propagates to all its subgroups.
Note that you need View Any Target privilege to see the full Administration Group hierarchy in the
Administration Group pages. Hence if you have a setup such as the one depicted in Figure 19 wherein
you only have privileges on a subset of groups in the hierarchy, then use the Groups page (accessible
from Targets → Groups) to view and perform operations on your groups.
Using Administration Groups for Other Group Operations

Administration Groups can be used in all features that support groups. This includes granting
privileges on these groups to your roles, using these as the targets for incident rule sets, blackouts,
system dashboard, etc. However, there can be scenarios where the membership criteria defined for
Administration Groups may not exactly match the desired membership criteria for these other group
operations. For example, you may have an Administration Group hierarchy based on Lifecycle Status
as shown below in Figure 22.
Figure 22. The Administration Groups in the hierarchy contain database and middleware targets.
The Administration Groups in the hierarchy contain database and middleware targets. For privilege
management purposes you may be unable to grant privileges on these groups to your roles because
you only want to grant the privileges on a subset of targets in the group. For example, you may want to
grant Full privileges on only the databases and not the middleware targets in the group to the DBA
team. One option would be to extend the Administration Group hierarchy by adding an additional
property to further breakdown the group (in our example, the new property would be Target Type).
This option however could make the hierarchy more complex especially if you already have a 3 or more
level hierarchy. The other option is to create dynamic groups that better satisfy the grouping
requirements. Dynamic groups are similar to Administration Groups in that dynamic groups are
defined by membership criteria. Once defined, Enterprise Manager automatically adds targets to the
appropriate dynamic group(s) if they match the dynamic group’s membership criteria. In our example,
you could create a dynamic group with membership criteria: Lifecycle Status = Mission Critical or
Production, Target Type = Oracle Database. Also set the Privilege Propagation attribute on the
dynamic group so that privileges granted on the group to a role or user automatically applies to all
members of the group. So for monitoring template apply purposes, Administration Groups would be
used; for privilege management purposes dynamic groups can be used. The two options are
summarized in Figure 23 below.
Figure 23. Options when Administration Groups do not meet other group requirements
In the end, you need to balance the tradeoffs between the options. Adding more levels to your
Administration Group hierarchy to accommodate non-monitoring requirements gives you the benefit
of being able to centrally manage all groups from a single group hierarchy structure but it adds more
complexity to your Administration Group hierarchy. Since the Administration Group hierarchy is
height-balanced, it could result in additional number of groups that will remain un-used. If you choose
to go the dynamic group route, then it doesn’t add unnecessary complexity to your Administration
Group hierarchy but it does require the management of a separate set of groups outside the
Administration Group hierarchy. While there is no one correct answer, the general recommendation is
to always create the Administration Group hierarchy based on monitoring requirements in order to
keep the hierarchy as simple and streamlined as possible. If there are different group membership
requirements for other non-monitoring needs, create dynamic groups to meet these other
requirements.
Verifying Targets are Part of the Administration Group Hierarchy
It is always good to check for orphaned targets, i.e. that there are no targets that should have been part
of an Administration Group but aren’t. Use the Unassigned Targets Report for this purpose (accessible
from the Associations page of Administration Groups).
Criteria for administration groups
Figure 24. The Unassigned Targets page helps you track targets that have been left out of the Administration Group
hierarchy.
The Unassigned Targets page shows you the targets that have not been added to any Administration
Group for any of these reasons:
1. The target property values do not match the Administration Group criteria. The report shows you
the target properties that are used for Administration Group criteria (represented by columns in the
table) as well as their current values. The top part of the page shows you the values of the target
properties used as criteria for the Administration Group hierarchy.
2. If the target is an aggregate target (i.e. a target that contain other targets as members), it is not
privilege propagating. This is indicated by the checkbox under the Non-Privilege Propagating
Aggregate column.
3. The target is still scheduled for membership evaluation into an Administration Group. The time in
which the target is scheduled for membership evaluation is shown in the ‘Pending Membership
Evaluation’ column. At that scheduled time, Enterprise Manager checks the target’s properties to
see if it matches the criteria of any Administration Group. If it does, it adds the target to the
appropriate Administration Group.
To address the first issue, set the target property values such that they match the Administration Group
criteria. You can set target properties via the Edit Target Properties page or via EM CLI
set_target_property_value. The target has to match all Administration Group criteria in order to join an
Administration Group. To address the second issue, set the aggregate target to be privilege
propagating via EM CLI modify_system with the –privilege_propagation=true option set.
Changing the Administration Group Hierarchy after Initial Creation
Even with careful planning of the Administration Group hierarchy, it is possible that you might have to
change it to accommodate new requirements. These changes include adding new target property
values (e.g. adding new Line of Business values which add more groups horizontally), merging two or
more groups, adding new levels, deleting levels, etc. It is possible to do these changes without re-
building the entire hierarchy. If you add a new level, that is the same as adding a new target property to
the Administration Group criteria. As such, you will need to set the value of this target property for all
your targets in order for them to continue to be part of the Administration Group hierarchy. When
deleting a level, that is the same as deleting a target property criterion. This causes groups at that level
to be deleted. If any of those groups have an associated template collection, then the monitoring
settings of the subgroups of the deleted group will be impacted since the subgroups obtained
monitoring settings from the associated template collection. You may need to review the remaining
template collections and re-associate the template collection with the appropriate Administration
Group.
If you want to merge two or more Administration Groups, this is done by merging their corresponding
target property criteria in the Administration Group hierarchy definition. The actual merge semantics
involve retaining one of the groups to be merged and moving over the targets from the other groups
into the group that is retained. Once the targets have been moved, the other groups will be deleted.
You get to choose the group to be retained. If the group has subgroups, then since the actual targets
reside in the leaf level (lowest level) Administration Groups, the actual movement of targets will occur in
the leaf level Administration Groups. The upper level Administration Groups will have updated criterion
based on the merged criteria.
As an example, consider the Administration Group hierarchy shown in the figure below. It is based on
Target Type and Lifecycle Status. The criteria associated with the Administration Groups are also
shown.
Figure 25. Administration Group hierarchy before the merge
Let us assume you want to merge the groups DB-Group and MW-Group. This is done by merging their
corresponding target property criteria ‘Database Instance’ and ‘Oracle WebLogic Server’.
Figure 26. Merging groups is done by merging their corresponding target property criteria.
In an Administration Group with multiple levels, each target property criterion corresponds to group(s)
containing that criterion. So, merging ‘Database Instance’ and ‘Oracle WebLogic Server’ target property
criteria means merging of DB-Group and MW-Group and their subgroups. As part of the merge, you
will be asked a question on which group to retain. Specifically, if you are merging ‘Database Instance’
and ‘Oracle WebLogic Server’, do you want to retain groups with criteria ‘Database Instance’ or groups
with criteria ‘Oracle WebLogic Server’? Assuming you choose to retain ‘Database Instance’, then the
following will occur:
• MW-Group will merge into DB-Group. This means the group called DB-Group will have its criteria
updated to include Oracle WebLogic Server.
• All member targets in Production group under MW-Group will move over to the Production group
under DB-Group. The Production group under DB-Group will now include ‘Oracle WebLogic
Server’ as part of its criteria.
• All member targets in Non-Production group under MW-Group will move over to the Non-
Production group under DB-Group. The Non-Production group under DB-Group will now include
‘Oracle WebLogic Server’ as part of its criteria.
• After the targets have been moved, the MW-Group, Production and Non-Production groups under
MW-Group will all be deleted.
The figure below shows a pictorial view of the merge process.
Figure 27. Merging involves moving over targets to the retained groups and updating its group criteria.
After the merge has been completed, the Administration Group now looks like this.
Figure 28. Updated Administration Group hierarchy after the merge
When choosing the group to retain, consider keeping the group that is used in most group operations
(e.g. incident rule sets, system dashboard, roles) so that the impact of the merge is minimized. The
operations on the retained group(s) will continue to remain intact and after the merge is done, the
operations will also apply on the targets newly added to the group.
Verifying Targets are in Sync with Your Monitoring Standards
As mentioned, templates are auto-applied to targets that are part of Administration Groups. In order to
verify whether or not targets in an Administration Group are in sync with their associated monitoring
templates, you can look at the Synchronization Status region in the group’s homepage:
Figure 29. The Synchronization Status region helps you verify if targets are in sync with the associated monitoring templates.
In the figure above, in the row Monitoring Template, the Synchronized Targets column shows the
number of targets that are in sync with the associated monitoring templates. This means the targets’
monitoring settings are the same as in the monitoring template 5. For this group, nine targets are in
sync with their associated monitoring templates. The next column, Pending Targets, shows the
number of targets that have pending template apply operations scheduled on them; hence their current
monitoring settings are not necessarily the same as the associated monitoring template. The date for
when this synchronization process will happen is indicated by the “Next Synchronization” date located
at the top of the table. You can wait for this sync to occur or perform the sync process now by clicking
on the ‘Start Synchronization” option at the top of the table.
The number under the Failed Targets column shows the number of targets on which the monitoring
template apply operation has failed. Finally, the N/A Targets show the number of targets on which
there is no associated monitoring template. A non-zero number under N/A Targets column is an
indication that you are missing some monitoring templates in your template collection. You can click
on the number under that column to find out the target types for which you have no monitoring
template included in your template collection.
If an administrator wants to check the sync status of all targets across all Administration Groups, go to
the homepage of the topmost Administration Group, and check its Synchronization Status region. It
should provide a rollup of the sync status across all targets in the Administration Group hierarchy.
For more details on the Synchronization Status region, refer to the “Implementing Administration
Groups and Template Collections” chapter of the Oracle Enterprise Manager Cloud Control
Administrator’s Guide 13c.
5The Synchronization Status region also shows whether or not targets are in sync with associated
compliance standards and cloud policies. These are, however, outside the scope of this paper.
Enabling Events for Jobs
The status of a job can change throughout its lifecycle – from the time it is submitted to the time it has
executed. For each of these job statuses, events can be raised to notify administrators of the status of
the job. By default, events are generated only for job status values that require administration
attention. These job status values include Action Required and Problem status values such as Failed or
Stopped. However, in order to avoid overloading the system with unnecessary events, job events are
not enabled for any target by default. Hence, if you would like to generate events for jobs, you will need
to specify the set of targets for which you would like job-related events to be generated. You can do
this using the Job Event Generation Criteria page which is accessible from the Setup → Incidents → Job
Events menu.
Integrating with Third Party Event Systems and Service Desks

Many datacenters often require integration with other management systems and/or service desk
systems as part of the monitoring solution. Enterprise Manager supports integration with other event
management systems to enable sharing of Enterprise Manager events with these other systems. This
type of integration can be done using Event Connectors. Once the appropriate event connector is
installed and configured, you can reference the event connector within your event rules. For more
details, refer to the event connector specific documentation that is available as part of the Enterprise
Manager documentation set.
The other type of integration is with service desk systems. Service Desk (or Helpdesk) connectors are
also available to enable the automatic opening of service desk tickets for events /incidents raised in
Enterprise Manager. Service desk specific documentation is available to guide you through the
installation and configuration of the connector. Once that is setup, then in your event or incident rules,
you can specify an action to open a service desk ticket for an event or an incident. The service desk
ticket ID and status are tracked and visible within Incident Manager. Thus, you have a central way to
monitor and track the status of all incidents, including those that are managed by service desk staff.
Too Many Alerts

Often you might think there are too many alerts (events) generated in your Enterprise Manager site. To
address these, think about controlling these events at the source. This means reviewing and
controlling the source of events rather than only focusing on clearing events after they have been
generated. Unnecessary events cause unnecessary load on your Enterprise Manager site. For metric
alerts specifically, it is important to review the target’s metric settings. To implement changes in metric
settings across your targets, you can make these metric setting changes in the monitoring templates
that you’ve included in your Template Collections:
• Disable metric collection for metrics you don’t care about

Most metrics for a target are enabled by default. If there are some metrics that you will not use
in your Enterprise Manager deployment, consider disabling the metric collection of these
metrics to avoid the unnecessary collection and storage of these metric values.
• Set thresholds only on metrics you care about
If your team typically ignores a set of metric alerts, then consider removing thresholds for
these metrics to avoid the generation of these metric alerts. You should only set thresholds
for metrics whose events (incidents) you will manage. You can still continue to collect metric
data values without alerting if the metric historical data is important for you (e.g. for reporting
or trending purposes).
• Adjust metric thresholds based on metric trend
Enterprise Manager 13c supports a threshold suggestion region in context of setting
thresholds for a metric. This is available as part of the Metric and Collection Settings page and
All Metrics page. This feature basically shows the value of the metric over recent periods of
time e.g. last 31 days:
Figure 30. The threshold suggestion region shows you the value of the metric over a recent period of time.
Use this region to review the historical value of the metric in order to fine-tune thresholds
defined for this metric. If, for a specific target, you end up deciding to use a threshold value
for the metric that is different from the associated monitoring template, remember to click the
‘Template Override’ checkbox for the metric in the Metric and Collection Settings page. This
ensures that any monitoring template applied to the target will not override your metric’s
threshold settings.
• Set number of occurrences
When a metric alert (event) should be generated only after a sustained period of time (e.g.
metric value is high for at least 30 minutes) use the number of occurrences parameter. If this
parameter is specified, it is used to determine the consecutive number of metric collection
intervals for which the metric must exceed its threshold before a metric alert (event) is
generated. If a metric is collected every 5 minutes, then if you specify a number of occurrences
parameter of 6, this means the metric needs to exceed its threshold for 30 minutes before a
metric alert (event) is generated (i.e. 5-minute collection interval multiplied by 6 number of
occurrences).
• Use corrective actions to auto-clear metrics alerts
If there are metric alerts that can be resolved in an automated manner, consider creating
Corrective Actions for these. Corrective Actions are job tasks associated with metrics such that
when an alert for the metric is generated, the corresponding corrective action (task) is
automatically executed. If the Corrective Action resolves the issue (e.g. restart the listener for
the listener down event), then the metric alert will be cleared by the agent during the metric’s
subsequent collection schedule. When setting up rules to send notifications or create
incidents for metric alerts, remember to consider any corrective action that might be in place
to avoid potential duplication of work between the recipient of the notification and the
corrective action.
• Avoid accumulation of resolved events by auto-clearing manually clearable events
There is a subset of events that require manual clearing by the administrator. Examples of
these are log-based events. As part of the workflow of managing the event, the administrator
should manually clear the incident that contains the event after the issue has been resolved.
(There is an option to clear an incident within Incident Manager). However, as a backup
option, you can consider creating a rule that will automatically clear these events after a period
of time. To avoid performance issues, this rule should be considered as a last resort to avoid
the accumulation of events and not used as the primary way to clear a large number of events.
To avoid prematurely clearing these events before someone has had a chance to look into
them, the period of time specified should cover the agreed-upon period of time by which these
events should have been resolved by the appropriate IT staff. For example, you might have an
operational procedure requiring these events be responded to and resolved within 14 days. In
that case, you can set up a rule that clears these manually clearable events only after 14 days.
Note that any rule you set up will apply to events that occur after the rule has been created, i.e.
the rule will not retroactively apply to events that already existed prior to the rule’s creation.
To clear these metric alert (events) that already exist, and as a general procedure to perform
bulk clearing of metric alert (events), use the EM CLI verb clear_stateless_alerts.
• Leverage the event compression functionality of incident rules to compress or combine related
events into a single incident.
Too Many Emails When a Host Goes Down

When a host goes down, all targets on the host go down as well. This will cause either Down or Agent
Unreachable events to be generated on the host, agent and all other targets on the host. In situations
such as these, it might be desirable to only notify the primary administrator(s) who need to address the
issue and avoid a flood of notifications to other administrators who own targets on the host server. The
system administrator is most likely the primary administrator who needs to address the problem at
hand by investigating and restarting the host. Any host start up script should also include the startup
of the agent to ensure monitoring is restored as soon as the host itself is back up. Do the following to
send notifications only to the designated system administrators in this situation.
• For system administrators who would like to know if a host goes down, create an event rule to create
an incident for selected ‘Target Availability’ events for the host target, specifically, the ‘Down’ and
‘Agent Unreachable’ availability states. If a host goes down and if its agent has a partner agent, then
an event will be raised on the host for the ‘down’ status. If there is no partner agent, then an event will
be raised on the host for the ‘Agent Unreachable’ status. Thus, choosing both states in the event rule
will cover both scenarios.
Figure 31. Choose ‘Down’ and ‘Agent Unreachable’ for host targets
For notification purposes, it is also recommended that you create an event rule for the ‘Up but
unmonitored’ state of the host and send notifications for these. This state occurs when the agent that
is monitoring the host is down but the host itself is still up. Since the agent is down, this also means
there is no monitoring of the host’s performance, space and other metrics.
• Create another rule to send notifications to the appropriate administrators for the host incident.
• Owners of the targets on the host will want to be notified if their targets are down (outside of the host
down scenario). For these administrators, create a rule to create an incident for the ‘Target
Availability’ event for these targets (database, listener, etc) and choose the ‘Down’ state. Also create a
rule to send notifications for this incident:
Figure 32. Choose 'Down' availability state for targets on the host.
With the rules set up in this manner, then if a host goes down, only the system administrators will
receive notifications for the ‘agent unreachable’ or ‘host down’ event/incident. All other target owners
will not be flooded with notifications because their targets are in ‘agent unreachable’ state and not
down state. Only if their targets are detected to be down (with the host up), will they be sent
notifications.
Using Root Cause Analysis for Target Down Events

When targets are detected to be down, root cause analysis for target down events will automatically
occur. This results in identifying whether a target down event is the root cause of other target down
events, or if the target down event is a symptom of other target down events, i.e. it has gone down
because the target on which it depends on has gone down. A new attribute called Causal Analysis
Update has been introduced to identify whether the event is a symptom or root cause. As additional
target down events come in, these events are analyzed and could result in updates to prior causal
analysis results. For example, a target down event that was neither root cause nor symptom could later
be identified as a root cause of another target down event when that second target down event is
reported 5 minutes later. Incidents containing these events will be classified accordingly, either as root
cause or symptom.
When root cause analysis is performed on target down events, the dependency relationship between
targets that are down is used in analyzing whether the target down event is a ‘root cause’ or ‘symptom’.
The causal analysis update attribute is used to show the results of this analysis. If the event is not
related to any other target down event, it will not have any value for the causal analysis update
attribute. The incident containing the target down event will have its causal analysis update attribute
reflect the same value as the causal analysis update attribute of the event.
You can use the root cause analysis feature in a several ways:
• Filter the System Dashboard and/or Incident Manager UI to exclude symptoms
In scenarios wherein a target down could in turn cause other targets to go down, you may want to
focus primarily on the ‘root cause’ event (or incident containing this) and exclude symptom events (or
incidents containing these). In the System Dashboard, you can specify an option to ‘Exclude
Symptoms’ from the Incidents and Problems table.
Figure 33. You can exclude symptoms from the Incidents and Problems table of the System Dashboard.
Similarly, in Incident Manager you can create a view and use the search field Causal Analysis Update
to exclude symptom events/incidents from your custom view.
Figure 34. You can filter out symptom events/incidents in your Incident Manager views.
• Create rules to create incidents only on non-symptom events
You may also want to leverage the root cause analysis feature in rules that create incidents.
Specifically, in scenarios where there could be a multitude of target down (symptom) events and only
a few (or single) target down event, you might want to create incidents and send notifications on only
the non-symptom events. To do this, create an event rule and select the Target Availability event
type and Target Availability State = Down. Also choose the ‘Causal analysis update’ option and select
‘event is marked as a cause’ and ‘event is not a cause and not a symptom’ as shown in the figure
below.
Figure 35. When creating an event rule, you can exclude symptoms as part of the rule criteria.
With the selections above, events that are marked as symptoms will be not be included in the rule. It
is possible that an event is initially reported as neither a root cause nor symptom. However, a few
minutes later, when other target down events are reported and analyzed, the original target down
event might later be classified as a symptom. Hence, you might consider introducing a small delay
(e.g. 5 to 10 minutes) in your rule to allow for causal analysis to have more recent updates. For
example, if your rule creates incidents based on non-symptom events, and you did not have a delay
in the rule, you could potentially have an incident created for an event that later turned out to be a
symptom. If the benefits of getting a potentially updated causal analysis outweigh the need to
perform an immediate action on an event (such as creating an incident or sending notification), add a
small waiting period in your rule by making the rule action a conditional action based on a duration.
This means you want the action in the rule to execute only if it meets a specified condition. In
general, the condition for an action can be based on criteria such as event severity or it can be based
on a length of time the event has been open. This latter option is what we want to use to introduce a
small waiting period in the event action. The figure below gives a specific example of how you could
configure the action in your rule with a small waiting period. In the figure below, we’ve selected the
option ‘Only execute the actions if specified conditions match’ and also ‘Event has been open for
specified duration’ and duration of 5 minutes. Introducing a 5-minute waiting period allows time for
other target down events to be reported and an updated causal analysis to occur before acting on the
event, i.e. before creating an incident for the event if it is not a symptom.
Figure 36. You can add a duration-based action with a small delay to allow for more updated causal analysis.
Managing Diagnostic Incidents and Problems

Most incidents created in Enterprise Manager can be considered operational in nature and are expected
to be resolved by a datacenter’s IT staff. There is, however, a special type of incident called a
‘Diagnostic Incident’ or ADR (Automatic Diagnostic Repository) Incident. These are incidents that are
automatically raised by Oracle software when it encounters a critical error in its software code. As such,
it is expected that IT staff reach out to Oracle (via Oracle Support) to help resolve these incidents. When
such incidents occur, an incident (object) and problem (object) are generated automatically in
Enterprise Manager. The incident represents an occurrence of the error and the problem (object) is
meant to help resolve these incidents by addressing the root cause of these incidents. The problem
object contains the incident’s “problem key”, i.e. something which uniquely identifies the software error.
Each occurrence of the incident will create a (diagnostic) incident that will automatically be associated
with the problem. To permanently resolve these incidents, the recommendation is to manage the
problem by using Support Workbench to package the appropriate diagnostic logs and open a Service
Request (SR) with Oracle Support. A link to Support Workbench is available in the Guided Resolution
section of the problem in Incident Manager:
Figure 37. A Problem (object) in Incident Manager enables management of critical errors in Oracle software.
The Support Workbench will assist in the creation of a draft SR including the packaging of the
appropriate diagnostic log information. Once a draft SR is created, you should complete the creation of
the SR within My Oracle Support. The SR will automatically be associated with the problem so that it
can be easily tracked with Incident Manager. The Problem object should be assigned an owner,
(resolution) status, etc. in the same way that other incidents are owned and managed in Incident
Manager. Once a fix to the problem has been provided (e.g. patch available for the software error),
after the fix has been applied, you can manually close the Problem in Incident Manager. This will
automatically close all its related incidents as well.
CONCLUSION
Comprehensive monitoring of applications and their supporting infrastructure continues to be a critical
requirement of today’s datacenters. To implement an effective monitoring solution, it is important for
IT staff to have the right management tool used in conjunction with best practice standards and
processes. It is also equally important to be able to automate IT processes to reduce administrative
overhead and enable IT staff to manage more with less while still providing high quality services to their
users. Oracle Enterprise Manager 13c provides a rich set of capabilities to meet the enterprise
monitoring demands of today’s dynamic datacenters and support and automate many of its
operational processes. The strategies outlined in this paper are designed to help IT staff plan and
optimize their use of Enterprise Manager 13c to meet their monitoring requirements in an effective and
scalable way.
CONNECT WITH US
Call +1.800.ORACLE1 or visit oracle.com.
Outside North America, find your local office at oracle.com/contact.
blogs.oracle.com facebook.com/oracle twitter.com/oracle
Copyright © 2020, Oracle and/or its affiliates. All rights reserved. This document is provided for information purposes only, and the contents hereof are subject to change without notice. This
document is not warranted to be error-free, nor subject to any other warranties or conditions, whether expressed orally or implied in law, including implied warranties and conditions of
merchantability or fitness for a particular purpose. We specifically disclaim any liability with respect to this document, and no contractual obligations are formed either directly or indirectly by this
document. This document may not be reproduced or transmitted in any form or by any means, electronic or mechanical, for any purpose, without our prior written permission.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and are trademarks or registered trademarks of SPARC International,
Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced Micro Devices. UNIX is a registered trademark of The Open Group. 0120
Strategies for Scalable, Smarter Monitoring Using Oracle Enterprise Manager Cloud Control 13c
May, 2020
Author: Ana McCollum
Contributing Authors: Karilyn Loui, H Siddaramaiah

Enterprise Manager 13 Monitoring Strategies

Uploaded by

Copyright:

Available Formats

Enterprise Manager 13 Monitoring Strategies

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Enterprise Manager 13 Monitoring Strategies

Uploaded by

Copyright:

Available Formats

What is the purpose of this document?

What is the purpose of this document?

What strategies are outlined to help optimize use of Enterprise Manager 13c?

What strategies are outlined to help optimize use of Enterprise Manager 13c?

Strategies for Scalable,

Smarter Monitoring using

April 2020 | Version 1.00

LAYING THE GROUNDWORK FOR MONITORING

Define and Set Up Standards for Monitoring

Organize Targets Based on How They are Monitored

Figure 1. Monitored targets further broken down by Line of Business

Defining Your Administration Group Hierarchy

Setting Target Property Values

Define Standards for Monitoring

Associate Template Collections with Administration Groups

Check if targets have

Defining Monitoring Templates for Primary and Standby Databases

Set the Synchronization Schedule

Define Roles for Different Job Responsibilities

Plan Job Responsibilities (Who can do what operations?)

Understand Enterprise Manager Privileges to support job responsibilities

Table 1. Enterprise Manager Privileges to Support Job Responsibilities

JOB RESPONSIBILITY ENTERPRISE MANAGER PRIVILEGE

Create Administration Group Full Any Target

Edit Administration Group hierarchy • Full Any Target

Delete Administration Group • Full Any Target

View entire Administration Group • View Any Target

Use Monitoring Templates • Create Monitoring Template resource privilege is required

Operations on the Administration

Operations on the members of the

Use roles to manage user privileges

Figure 13. This shows the privilege-propagating nature of Administration Groups.

Create Roles for different job responsibilities

JOB RESPONSIBILITY ROLE IN ENTERPRISE *PRIVILEGES IN THE ROLE

Group Administrator GROUP_ADMIN_ROLE Group Administration on the group

Senior Administrator SENIOR_ADMIN_ROLE • Add Any Target

Target Owner TARGET_OWNER_ROLE • Operator on the Administration

First Level Support FIRST_LEVEL_SUPPORT • Manage Target Events on the

Assign Roles to your Enterprise Manager administrators

Set up Rules to Manage Events and Incidents

• Understanding rule sets and rules

• Using groups in rule sets

• Auto-creating incidents using rules

• Planning your rules to leverage Enterprise Manager rule features

• Implementing the rules in Enterprise Manager

Events and Incidents

Figure 14. This is an example of an incident containing one event.

Figure 15. This is an example of an incident containing two events.

Understand Rule Sets and Rules

Use Groups in Rule Sets

Auto-create Incidents Using Rules

Minimizing Event Storms by Compressing Multiple Events into a Single Incident

Plan Your Rules

Rules for the production group PROD:

Which type of rule should I use?

TYPE OF RULE PURPOSE

Rule on Events • Create incidents based on one or more events

Rules on Incidents • Automate incident workflow operations (e.g. assignment of

Rules on Problems • Automate problem workflow operations (e.g. assignment,

Rule Set name: Rule set for production group PROD