XCON: General Data Gathering Procedures for MTA Troubleshooting

ID: Q198688

The information in this article applies to:

Microsoft Exchange Server, versions 4.0, 5.0, 5.5

IMPORTANT: This article contains information about editing the registry. Before you edit the registry, make sure you understand how to restore it if a problem occurs. For information about how to do this, view the "Restoring the Registry" Help topic in Regedit.exe or the "Restoring a Registry Key" Help topic in Regedt32.exe.

SUMMARY

This article is designed to help Microsoft Exchange Server administrators troubleshoot message transfer agent (MTA) problems on one or more Microsoft Exchange Server computers. Although this article discusses MTA issues, other Microsoft Exchange Server Support Professionals often request similar data. Furthermore, you may require different sets of data to diagnose different MTA problems. For example, to troubleshoot MTA content conversion problems you may require specific data that you may not require to troubleshoot a routing problem.

The data gathering techniques in this article help you diagnose Exchange Server problems, and also help when you call Microsoft Product Support Services (PSS). Before you call Microsoft PSS, be sure you gather the system and application event logs from the problem server or servers, and have them packaged and ready to send to Microsoft Support Professionals, either in an e-mail or using File Transfer Protocol (FTP). This is the minimum amount of information you need to expedite case resolution. The scope of Exchange Server MTA problems includes connector problems (Dynamic RAS, Site, or X.400), high CPU use by the MTA, MTA not starting or running incorrectly, and so on.

MORE INFORMATION

Many Exchange Server administrators are not aware of the type or scope of information that Microsoft Support Professionals need from them. When you open a support issue regarding an MTA problem, Microsoft recommends that you have the following elements on hand at the time that you contact Microsoft PSS. You may need to collect additional data, but the following guidelines provide a good baseline data set for identifying and resolving MTA problems. This type of data collection is sometimes difficult to coordinate and is not always possible, but it is very helpful in isolating a problem.

System and Application Event Logs

The system event log is necessary for standard troubleshooting procedure. It can contain information about problems (with NTLM authentication, name resolution problems, and so on) that may actually be causing the Exchange Server problem.

Application event logs are extremely helpful in diagnosing single server MTA problems, or connection related failures between servers. As a general rule, when MTA problems arise there are at least three categories of application event logging detail that you want on the MTA - X.400 service, interface, and internal processing. Increase these three categories to maximum logging levels, increase the application event log size to 20 to 30 megabytes (MB), and in the Event Viewer, on the Options menu, click Overwrite as Needed. You must clear the application event log for a log size increase to take effect.

For additional information about clearing the application event log, please see the following article in the Microsoft Knowledge Base:

Q142032 PRB: Event Log Does Not Resize Without Clearing

Be sure to save the file with the .evt extension.

At the time of this writing, Microsoft has an incoming size limit of 5 MB on e-mail messages, so before you send logs close to or over this size limit in e-mail to a Microsoft Support Professional, zip them. If the zipped files are still too large, you can send them using FTP. Ask your Microsoft Support Professional for FTP instructions.

Obtain system and application event logs at each affected MTA; this is important because the MTA communication is an end-to-end process, and analyzing logs from just one of the affected MTAs may not be adequate to determine the problem. Errors may appear in one MTA log that indicate a connectivity problem, but the opposing MTA may itself be the real problem. For this reason, you need logs from each server affected by the problem.

For additional information about logging options for the MTA, please see the following articles in the Microsoft Knowledge Base:

Q153188 Description of MTA Diagnostics Logging Options

Q168906 Setting up Advanced Logging on Exchange 5.0 and 5.5 MTAs

Q163033 Description of Diagnostics Categories for MTA

Performance Monitor Logging

Performance Monitor logging for all objects on the server is useful in determining not only the state of the Exchange Server MTA, but for all underlying system processes, including, but not limited to, other Exchange Server services. Logging MTA objects (MSExchangeMTA and MSExchangeMTA Connections) is important for MTA problems, but if all objects are logged you can see if another process is contributing to MTA problems.

To gather Performance Monitor log files:

Start Performance Monitor (not on the server you want the Performance Monitor log for), and on the View menu, click Log.

On the Edit menu, click Add to Log, select the Exchange Server computer you want a Performance Monitor log for, and then select all objects. Click Add, and then click Done.

On the Options menu, click Log, and set the interval to 30 seconds. Name the log Servername.log.

Click Start Log, and let logging run for 20 to 30 minutes. Then, on the Options menu, click Log, and then click Stop Log.

Send Microsoft PSS the Performance Monitor log file.

For additional information about how to create a Performance Monitor log, please see the following article in the Microsoft Knowledge Base:

Q150934 How to Create a Performance Monitor Log for NT Troubleshooting

If possible, generate a Performance Monitor log for each server, sampling ALL objects every 30 seconds, for at least 20 to 30 minutes, to give Microsoft PSS a "snapshot" of the overall server conditions at the time. Be aware that if you do not monitor the log, it increases in size and eventually fills your hard disk drive. If you monitor a server for an extended period of time, increase the update interval. You can also take baseline Performance Monitor statistics during a non-problem time period, to compare those statistics with statistics from the time of the issue.

Winmsd Reports from All Affected Servers

This report gives useful details about the server's services, drivers, environment, and so on. It is a relatively small file that is easy to create, and it can help you troubleshoot problems such as an unresponsive server. To gather a Winmsd report:

At the server you want a Winmsd report for, click the Start button, point to Run, type winmsd, and then press ENTER. Do this at the problem server, because when you do it remotely, some information, such as the Drives report, may be omitted.

On the File menu, click Save Report.

Under Scope, select all tabs.

Under Detail Level, select Complete.

Under Destination, select File.

Save the file as Winmsdservername.txt

A Network Trace

A trace captured by Microsoft Network Monitor, or an equivalent program, is also helpful in analyzing packet level data and determining if a connection-related problem is a result of some network-level problem. Microsoft Support Professionals prefer data captured with Network Monitor, but data captured with the Network General Sniffer (.enc or .trc files) can be converted to Netmon (.cap) format.

You must capture this trace at the same time as the event logs noted above to correlate the connection failures to the data in the trace. To minimize the size of the trace file, capture all traffic to and from the media access control (MAC) or IP addresses of the problem server or servers.

Set the buffer size of the capture file to at least 10 MB. For best results, set a trigger to stop the capture when 100 percent of the buffer is filled. This protects critical capture data from overwriting.

You must capture the trace on the same physical network segment as that of the problem server, either from the server itself or from a workstation. To capture remotely you need the full Microsoft Systems Management Server version of Network Monitor, not the standard version that comes with Windows NT.

NOTE

For additional information about capturing and interpreting network traffic with Network Monitor, please see the following articles in the Microsoft Knowledge Base:

Q148942 How to Capture Network Traffic with Network Monitor

Q159298 XADM: Analyzing Exchange RPC Traffic Over TCP/IP

Q169292 The Basics of Reading TCP/IP Traces

A Calls.out File

The Calls.out file is a text file that details the state of the MTA internal processes. A Calls.out file is useful when the MTA does not respond to a Stop command, or when the MTA runs but does not process mail. A Calls.out file takes little time to create.

WARNING: Using Registry Editor incorrectly can cause serious problems that may require you to reinstall your operating system. Microsoft cannot guarantee that problems resulting from the incorrect use of Registry Editor can be solved. Use Registry Editor at your own risk.

For information about how to edit the registry, view the "Changing Keys and Values" Help topic in Registry Editor (Regedit.exe) or the "Add and Delete Information in the Registry" and "Edit Registry Data" Help topics in Regedt32.exe. Note that you should back up the registry before you edit it. If you are running Windows NT, you should also update your Emergency Repair Disk (ERD).

Access the registry a the following location:
HKLM/System/CCS/Services/MSExchangeMTA/Diagnostics
Select any one of the diagnostic logging categories.

Set the value for that category to 7, wait for approximately 5 seconds, and then set the value back to its previous setting.

This generates a file named Calls.out in the Mtadata directory on the server.

Rename and back up this file immediately, as it is automatically deleted by the MTA when you restart services. Create a Calls.out file at each server that is experiencing the problem.

The Calls.out file is most useful if the server is running a debug version of the MTA, but the file can also contain valuable information if it is created with a non-debug version.

General IP and Network Information

At a command prompt, type ipconfig -all>c:\ipconfig.txt. This creates a text file that you can use to verify IP addresses, host names, node type, and so on. This information is useful in diagnosing connector and routing problems.

Routing Table Information

For routing and delivery problems, you may need to view the Gwart0.mta file, found by default in the Exchsrvr\Mtadata directory. This file contains a representation of all the routes known by the MTA that are necessary for mail delivery.

For additional information about the Gateway Address Routing Table, please see the following article in the Microsoft Knowledge Base:

Q149121 Gateway Address Routing Table Information

Creating an Admindmp.txt File on an Exchange Server Object

Instead of checking each property page of an object in the Exchange Server directory and dictating the information over the phone, it is easier, faster, and more accurate to create a file that includes those details and send it to Microsoft PSS. For example, if an X.400 Connector is not working between two Exchange Server computers, a Microsoft Support Professional can look at these files (one created at each end of the connector) to find the problem.

WARNING: Using the raw mode of the Exchange Server Administrator program (admin /r) incorrectly can cause serious problems that may require you to reinstall Microsoft Windows NT Server and/or Microsoft Exchange Server. Microsoft cannot guarantee that problems resulting from the incorrect use of raw mode can be solved. Use raw mode at your own risk.

Before you perform this procedure, be sure there is not a file named Admindmp.txt in the Exchsrvr\Bin directory. If the file exists, move, rename, or delete it. Each time you create this file, it appends to the most recent existing file of the same name. To simplify your analysis, you need to create a new Admindmp.txt file.

Start the Exchange Server Administrator program in raw mode, change to the Exchsrvr\Bin directory, and at the command prompt, type admin -r.

Select the object you want to create an Admindmp.txt file for, and press and hold down the CTRL key on your keyboard. Using your mouse, on the File menu, click Raw Properties.

Hold down the CTRL key until the Properties window appears. After the Properties window appears, you can close it.

A file named Admindmp.txt is created in the Exchsrvr\Bin directory. Rename the file Servername.txt.

If the object in question is a connector, repeat this procedure at the other end of the connector.

For additional information about creating an Admindmp.txt file, please see the following article in the Microsoft Knowledge Base:

Q146468 XADM: Creating Admindmp.txt to Log Objects Raw Properties

Additional query words: Perfmon

Keywords : exc4 exc5 exc55
Version : winnt:4.0,5.0,5.5
Platform : winnt
Issue type : kbhowto