Exchange Management Pack Monitoring Scenarios


Topic Last Modified: 2006-03-28

There are four potential scenarios in which you can monitor an Exchange Server 2003 organization with the Exchange Management Pack:

  • An agentless environment

  • A centralized environment

  • A distributed environment

  • A hybrid environment

Each scenario is discussed in the following sections.

In an agentless environment, no agents are installed on the Exchange servers being monitored by MOM. MOM automatically monitors your servers without first installing an agent. The Exchange Management Pack will monitor performance counter and event-related problems in this environment, but cannot run scripts.

If you plan to monitor servers that do not have agents installed, you must be aware of the limitations of this configuration. By default, the Exchange Management Pack will only monitor servers on which the MOM agent is installed. If you want to monitor agentless servers, you must manually add each agentless server to the appropriate computer group. The computer groups are as follows:

  • Microsoft Exchange 2000 Server Back-end

  • Microsoft Exchange 2000 Server Front-end

  • Microsoft Exchange Server 2003 Back-end

  • Microsoft Exchange Server 2003 Front-end

To add a server to a computer group
  1. Click Start, point to Programs, point to Microsoft Operations Manager 2005, and then click Administrator Console.

  2. In the MOM 2005 Administrator Console, in the Console Root, expand Microsoft Operations Manager\Management Packs, and then click Computer Groups.

  3. In the right pane, right-click the computer group that you want to add a server to, and then click Properties.

  4. In the Properties dialog box, click Included Computers.

  5. On the Included Computers tab, add the agentless server to the group, and then click OK to close the dialog boxes.

The Exchange Management Pack for MOM 2005 supports the following monitoring features on agent-managed computers:

  • Event monitoring

  • State monitoring

  • Service discovery

  • Service availability

  • Database health

  • Disk space monitoring

  • Health and availability monitoring

  • Database configuration monitoring

  • Service pack compliance

  • Server performance threshold monitoring

  • Event collection

  • Server performance collection

  • Client monitoring

This management pack supports the following features on Exchange Server computers that are monitored without an agent:

  • Event monitoring

  • Server performance threshold monitoring

  • Event collection

  • Server performance collection

Note that the MOM service account on the MOM server requires administrator access to agentless computers, but the MOM service account on a managed computer does not. The MOM Agent Action account on each Exchange server should be running as Local System.

To update the Agent Action Account
  1. In the MOM 2005 Administrator Console, expand Microsoft Operations Manager\Administration\Computers, and then click Agent-managed Computers.

  2. In the left pane, right-click the server that you want to update, and then click Update Agent Settings.

  3. In the Update Agent Settings Task dialog box, under Which account do you want to use for the Agent Action Account, click Local System, and then click OK.

In a centralized environment, a single location contains servers, routing groups, and administrative groups. This helps reduce administrative overhead, and also enables convenient monitoring. For example, a centralized environment can have a front-end and back-end architecture with users in the central location who have MAPI access for their Outlook clients, and remote users who use Outlook Web Access and connect through the front-end servers.

For centralized environments with up to 2000 servers, the single management group topology provides a fitting, centralized administrative model. This MOM topology can monitor one or more domains and contains one or more MOM Management servers. For enterprises that are monitoring fewer than 50 computers, a single configuration group topology can be used with a single computer that is running all MOM components. For larger organizations, MOM components can be put on different computers, as shown in this MOM single configuration group figure.

MOM single configuration group

MOM single configuration group

A distributed environment extends the centralized environment concept by giving locations control over their systems. In a distributed environment, routing groups and administrative groups are monitored from various locations. An example of this is a corporation after a merger. The separate messaging systems might require separate administration and monitoring, in addition to established routes for message flow, or a merged system might require access by two different departments.

Multiple configuration groups enable monitoring of multiple locations in a distributed environment. To allow for uniform data gathering, MOM has a feature named multi-homing that lets you configure agents as members of more than one configuration group. For example, many enterprises consist of organizations with multiple departments that are responsible for managing different aspects of a server. One group might monitor security issues, and another group might monitor only servers that are running specific Windows applications, such as Exchange.

In a multi-homed agent environment, the agent reports to multiple configuration groups for the events, alerts, and performance counters that it collects. The multi-homed agent processes work from the different configuration groups independently of one another so there is no conflict of rules. Agent computers can belong to up to four configuration groups. You can deploy this topology across one or more domains. The following figure shows a distributed environment monitoring scenario.

MOM multiple configuration groups with multi-homed agents

MOM multiple configuration groups

A hybrid environment combines centralized and distributed environments. For example, suppose you require all mailboxes to be in a central location, and you also want to let other locations have control over different server resources, such as public folders. This is an example of a hybrid approach that combines both centralized and distributed environments.

Monitoring in this scenario is similar to monitoring a centralized environment. Tiers of configuration groups and alert forwarding can be used to monitor multiple physical locations from a central point. Alert forwarding makes this monitoring easier by creating a hierarchical monitoring infrastructure.

Alert forwarding is used to achieve centralized monitoring by forwarding only alerts and the events that are associated with those alerts. Alert forwarding enables consolidators in one configuration group to send alerts to another configuration group, which creates an efficient hierarchical alert-management structure for large enterprise networks, and can reduce network bandwidth requirements.

Establishing a master configuration group to receive alerts, and then establishing one or more zone configuration groups to send alerts accomplishes alert forwarding. The alerts and associated events are kept separately in the zone configuration group database and in the master configuration group database, so the data in the zone configuration group can be used for trend analysis.

Alerts that are forwarded maintain the name of the source computer where the actual event was generated. Alert responses are processed independently between the zone and master configuration groups. Changes that are made in the alert properties in a zone configuration group do not affect alerts in the master configuration group and vice versa.

MOM supports a two-tiered architecture: zone configuration group to master configuration group. Implementing more than two tiers is not supported. A master configuration group can support up to ten zone configuration groups that have up to 120,000 alerts per day forwarded to the master configuration group.

Designing and operating a multi-tiered topology requires significant team effort and coordination to implement processing-rule changes and configuration changes.

The following figure shows the topology for monitoring a hybrid environment.

Multi-tiered configuration groups with alert forwarding

Multi-tiered configuration groups w/alert forward

Community Additions