What are the Outlook Performance Counters used for?

Microsoft Office Outlook performance monitor objects can be viewed, logged, and archived just like any other counter using Performance Monitor.

William Lefkovics

December 22, 2008

4 Min Read
ITPro Today logo

When you install Microsoft Office 2007 (except Microsoft Home and Student version) or just Microsoft Office Outlook 2007, several performance counters get added for monitoring Outlook 2007 connectivity. These counters were introduced with Microsoft Office Outlook 2003. The performance monitor objects can be viewed, logged, and archived just like any other counter using Performance Monitor.

Windows ships with a monitoring tool called Performance Monitor. It doesn’t collect performance data by default; it needs to be configured. You can start Performance Monitor on all recent versions of Windows by going to Start, Run and entering Perfmon. Windows Vista will show the Reliability and Performance Monitor window, and Windows XP will show the Performance window. In Windows Vista select Performance Monitor in the left pane to view Perfmon. When Perfmon is started, by default there are a few standard counters running. To add counters, click the “+” button, or right-click in the output window and select Add Counters from the context menu. The Add Counters window displays the list of available counters for the workstation. Here you can select Outlook from the drop-down box for Performance Objects. After you’ve made your selection, the different counters available will be listed below it. Figure 1 shows the list of counters in Windows XP, and Figure 2 shows the list of counters in for Vista.

There are thirteen counters which can be separated into three main categories of performance counters available for Outlook.These counters are the same for Outlook 2007 and Outlook 2003.

Connections

Count obj connection—The number of connection objects that are currently being used.

Number of RPCs

  • RPCs Attempted—The number of remote procedure calls (RPCs) that Outlook tried to send to the server.

  • RPCs Attempted – UI—The number of RPCs that Outlook tried, but they were blocked by the UI.

  • RPCs Canceled—The number of RPCs that were sent to the server, but that the user canceled.

  • RPCs Failed—The number of RPCs that were tried and that failed. RPCs Succeeded—The number of RPCs that Outlook successfully sent to the server.

  • RPCs UI Shown—The number of RPCs that were sent to the server, and that took long enough to show progress UI.

RPC performance (and latency)

  • Time Avg (10)—The average time (ms) that it took for the last 10 RPCs to complete successfully.

  • Time Avg (50)—The average time (ms) that it took for the last 50 RPCs to complete successfully.

  • Time Avg (200)—The average time (ms) that it took for the last 200 RPCs to complete successfully.

  • Time Avg (all)—The average time (ms) that it took for all RPCs to complete successfully.

  • Time Max—The maximum time (ms) that it took for an RPC to complete successfully.

  • Time Min—The minimum time (ms) that it took for an RPC to complete successfully.

The Count Obj Connection shows the number of items that Outlook is connecting to. This number goes up as the user opens different components of his or her own mailbox as well as other shared items in the store, such as a shared calendar from another mailbox. The number of RPC counters quantify the number of RPCs between the client and the Exchange Server. These also increase with MAPI client activity. RPCs are a critical component for MAPI communication. There should not be many failed RPCs. RPC Performance charts the time it takes for RPC calls between Outlook and Exchange Server. A lot of these counters are the same but are of different scale. The smaller the values for RPC communication time the better, but charting an average over time can identify trends for RPC performance. Normal ranges in this case are dependent on the activity level of the Exchange Information Store. Fluctuations in average RPC time for Outlook are specific to use. For example, some MAPI events, such as a mailbox search, are expected to take longer than others. This can skew RPC latency times in the short term. In addition, the number of RPCs is lower when Cached Mode is in use. Figure 3 shows the Outlook counters being displayed in Performance Monitor on Windows XP.

Performance monitor counters for Outlook are beneficial only for gauging or troubleshooting a single Outlook installation, although comparing performance metrics between similar workstations or against a benchmark can help identify problems as well. The Exchange Server itself has a wide array of RPC monitoring features, including performance counters, to identify issues at the server level.

Sign up for the ITPro Today newsletter
Stay on top of the IT universe with commentary, news analysis, how-to's, and tips delivered to your inbox daily.

You May Also Like