Monitoring Server Health and Performance

Good server health is important to a stable and reliable messaging environment. Using a variety of tools included with Windows 2003 and Exchange 2003, administrators can maximize the uptime of the messaging environment by monitoring items such as memory usage, CPU utilization, SMTP queues, or free disk space.

Using the Performance Console

The Performance Monitor Console (PerfMon) is a toolbox that provides many ways to monitor and analyze server performance issues, network bottlenecks, or gather baselines of server usage. When using Performance Monitor, be methodical and always check these three Windows server resource objects: memory, processor, and disk. After you have checked that they are not causing the bottleneck, check Exchange-specific countersfor example, MSExchangeIS (Send and Receive queues size) and MSExchangeMTA (MSExchangeMTAWork Queue Length). Some benefits of using PerfMon are understanding what components are doing (for example, is the disk reading or writing most of the time?). Information gathered can be displayed in different formats such as reports and real-time charts or logs, and can help you to know where a minimum investment will produce the maximum gains (such as adding RAM). PerfMon, as shown in Figure 12.2, is launched by selecting Start, All Programs or Programs, Administrative Tools, Performance.

Figure 12.2. Viewing server performance in the Performance Monitor Console.

Using Network Monitor

The Network Monitor, as shown in Figure 12.3, is tool that specifically monitors network traffic. Two versions of Network Monitor ship with Microsoft products. The version that comes with Windows Server 2003 is the watered-down version and only allows analysis of traffic sent to or from the server that Network Monitor is running on. The full version of Network Monitor, included with SMS Server, allows monitoring of any machine on your network, determining users' bandwidth consumption and protocol bandwidth consumption, and more. Network Monitor is launched by selecting Start, All Programs or Programs, Administrative Tools, Network Monitor. If prompted, select the network connection to monitor, and then select Capture and click Start.

Figure 12.3. Viewing network traffic via Network Monitor.

Tip

Network Monitor (NetMon) is not installed by default with Windows 2003, but adding it is easy following these steps:

1.	On the Start menu, select Control Panel and then select Add or Remove Programs.
2.	Select Add/Remove Windows Components and then choose Management and Monitoring Tools. Click the Details button, select the Network Monitor Tools check box, and then click OK.
3.	Follow the instructions until the installation is complete. Administrators can then find NetMon in the Administrative Tools folder (Start Menu, Administrative Tools, Network Monitor).

Using Task Manager

Administrators can get a quick overall idea of how Exchange 2003 is performing at any given time by launching Task Manager: right-click the taskbar and then click Task Manager. Along the top of the Task Manager window are tabs for Applications, Processes, Performance, Networking, and Users. Each tab provides a brief summary of information. The last two tabs were added because of customer requests for information about networking and user-specific attributes. One caveat with Task Manager is that it cannot save historical datadata that is useful for setting baselines and capacity planning.

Monitoring Processor Usage

When CPU utilization is high, Exchange performance suffers greatly. CPU utilization of 100% can be indicative of serious server problems. If a server is stuck at 100% CPU utilization, rebooting is usually necessary to return the server to a functioning level. To keep a check on CPU utilization, configure CPU monitors using the following steps:

1.	In ESM, expand the administrative group of the server to work with. Expand Servers, right-click the server to work with, and then select Properties.
2.	On the Monitoring tab, click Add. In the Add Resource dialog box, select CPU Utilization and then click OK, displaying the CPU Utilization Thresholds dialog box.
3.	Type the number of minutes that must be exceeded to trigger the state change in the Duration (minutes) field. Typically, a value between 5 and 10 minutes should be entered.
4.	Configure entries in the Set Maximum CPU Utilization Thresh olds section. Select the check box next to Warning State (percent) or Critical State (percent) and then enter values into the associated fields. Typically, a value of 90% is entered for the Warning State threshold and a value of 100% is entered for Critical State threshold.
5.	Click OK to finish configuration. If desired, use the following tip to configure automated administrator notification.

Tip

In ESM, right-click the Monitoring and Status folder, click the Notifications subfolder, and select New, Email Notification to create an email notification using built-in scripts. Exchange provides proformas with variables that are substituted with figures at runtime. But waitisn't the server down? The work around is to monitor one server from another server. Choose users with mailboxes on different servers so that when a service fails, at least one email notification should get through to an administrator's inbox.

Monitoring Disk Subsystems

In Exchange, free disk space is used for tracking messages, logging, data storage, and virtual memory. Running out of hard disk space will cause data loss and Exchange Server errors. Free disk space should be monitored closely. Configure disk monitoring using the following steps:

1.	In ESM, expand the administrative group of the server you want to work with. Expand Servers, right-click the server to work with, and then select Properties.
2.	On the Monitoring tab, click Add. In the Add Resource dialog box, select Disk Free Space and then click OK, displaying the Disk Space Thresholds dialog box.
3.	Select the drive to monitor, such as C:, in the Drive to Be Monitored drop-down list.
4.	Configure entries in the Set Maximum Drive Space Thresholds section. Select the check box next to Warning State (MB) or Critical State (MB) and then enter values in the associated fields. Typically, a value of 500MB is entered for the Warning State threshold and a value of 100MB is entered for Critical State threshold.
5.	Click OK to finish configuration. If desired, use the preceding tip to configure automated administrator notification.