Chapter 21 Managing Logging

This chapter provides overview information on the logging facilities for the Messaging Server MTA, the Message Store, and services. This chapter also provides procedures for how to manage these logging facilities.

This chapter contains the following sections:

• Overview of Logging

• Tools for Managing Logging

• Managing MTA Message and Connection Logs

• Managing Message Store, Admin, and Default Service Logs

Overview of Logging

Logging is the means by which a system provides you with time-stamped and labeled information about the system’s services. Logging provides both a current snapshot of the system as well as a historical view.

By understanding and using Messaging Server log files, you can:

• Gather message statistics, such as message size, rate of message delivery, and how many messages are passing through the MTA

• Perform trend determination

• Correlate capacity planning

• Troubleshoot problems

For example, if your site needs to add more disk storage due to an increase in the number of users, you can use the Messaging Server log files to see what percentage your system demand has increased by and plan for the amount of new disk storage you need.

You can also use Messaging Server logs to understand what your messaging pattern looks like across one day. Understanding when your daily peak loads occur helps you conduct capacity planning.

Logging is also helpful for troubleshooting user problems. For example, if a user isn’t receiving expected mail messages, you can use the Messaging Server logging facilities to trace the user’s mail messages. In so doing, you might find out that the messages didn’t arrive because they were automatically filtered and sent to a SPAM folder.

Types of Logging Data

In general, logging provides you with two types of information:

• Operational data

• Error conditions, also known as event logging

For the most part, Messaging Server logging provides operational data. This operational data contains information such as: the date and time a message entered the system; the sender and recipient of the message; when the message was written to disk; and at a later point in time, when the message was removed from disk and inserted into user’s mailbox.

However, Messaging Server logging does also provide some event logging data. To obtain event logging data, you need to pull together multiple items from different log files. You could then use a unique constant, such as message ID, to search and correlate the lifecycle of a message as it passed from point to point through the system.

Types of Messaging Server Log Files

Messaging Server logging consists of three types of log files:

1. MTA logs. These logs provide operational data previously described for the Message Transfer Agent.

2. Error logs. These are the MTA debug logs, and the MTA subcomponent logs (that is, job controller, dispatcher and so on).

3. Message Store and Service logs. These logs provide messages from the http server, mshttpd, imap, and pop services, as well as the Admin service. The format of these logs differs from that of the first two types of logs.

The following table lists the different types of log files. By default, log files are located in the msg_svr_base/data/log directory. You can customize and view each type of log file individually.

where:

sequenceNum - specifies an integer that specifies the order of creation of this log file compared to others in the log-file directory. Log files with higher sequence numbers are more recent than those with lower numbers. Sequence numbers do not roll over; they increase monotonically for the life of the server (beginning at server installation).

timeStamp - Specifies a large integer that specifies the date and time of file creation. (Its value is expressed in standard UNIX time: the number of seconds since midnight January 1, 1970.)

For example, a log file named imap.63.915107696 would be the 63rd log file created in the directory of IMAP log files, created at 12:34:56 PM on December 31, 1998.

The combination of open-ended sequence numbering with a timestamp gives you more flexibility in rotating, expiring, and selecting files for analyzing. For more specific suggestions, see Defining and Setting Service Logging Options

Tracking a Message Across the Various Log Files

The following describes how a message flows through the system, and at what point information gets written to the various log files. This description is meant to aid you in your understanding of how to use Message Server’s log files to troubleshoot and resolve problems. See Figure 8–2 to follow along.

1. A remote host makes a connection to the TCP socket on your messaging host, requesting SMTP service.

2. The MTA dispatcher responds to the request, and hands off the connection to your messaging host’s SMTP service.

   As the MTA is modular in design, it consists of a set of processes, including the job controller and the SMTP service dispatcher. The dispatcher takes the incoming TCP connection and sends it to the SMTP service. The SMTP service writes the message to disk to a channel area. The SMTP service understands the message’s envelope parameters, such as sender and recipient. Configuration entries in the system tell what destination channel it belongs to.

3. The dispatcher writes to the dispatcher.log file that it forked a thread and made the thread available to incoming connection from a certain IP address.

4. The SMTP server writes to its tcp_smtp_server.log file, recording the dialog of what happens when the remote host connected to it and sent a message. This log file gets created when dispatcher hands off to SMTP server on the host’s IP.

5. The SMTP server writes the message to a queue area on disk for a channel program such as tcp_intranet, and informs the job controller.

6. The job controller contacts the channel program.

7. The channel program delivers the message.

   Each channel has its own log file. However, these logs usually show the starting and stopping of the channel. To get more information, you need to enable debug level for the channel. However, as this can slow down your system and actually make problems more obscure if left on, you should only enable debug level when an actual problem is occurring.

   ---

   Note –

   For efficiency, if a channel is already running for an existing process, and a new message comes in, the system does not spawn a new channel process. The currently running process picks up the new message.

   ---

8. The message is delivered to its next hop, which could be another host, another TCP connection, and so forth. This information is written in the connection.log file.

   At the same time that the SMTP server writes the message to a queue area on disk, the channel responsible for the message writes a record in the mail.log_current, or mail.log file. The record shows such information as the date and time the message was enqueued, the sender, the recipient, so forth. See MTA Message Logging Examples for more information. The most useful file for tracing the message is the mail.log_current file.

Tools for Managing Logging

You can customize the policies for creating and managing Messaging Server log files by using Console and the configutil command.

For Message Store logs, you can use Console to specify log settings and to view logs. The settings you specify affect which and how many events are logged. You can use those settings and other characteristics to refine searches for logged events when you are analyzing log files.

Because the MTA uses a separate logging facility, you cannot use Console to configure MTA logging services and view logs. Instead, you configure MTA logging by specifying information in configuration files.

For log analyses and report generation beyond the capabilities of Messaging Server, you need to use other tools. You can manipulate log files on your own with text editors or standard system tools.

With a scriptable text editor supporting regular-expression parsing, you can potentially search for and extract log entries based on any of the criteria discussed in this chapter, and possibly sort the results or even generate sums or other statistics.

In UNIX environments you might also be able to modify and use existing report-generation tools that were developed to manipulate UNIX syslog files. If you wish to use a public-domain syslog manipulation tool, remember that you might need to modify it to account for the different date/time format and for the two extra components (facility and logLevel) that appear in Messaging Server log entries but not in syslog entries.

Managing MTA Message and Connection Logs

The MTA provides facilities for logging each message as it is enqueued and dequeued. It also provides dispatcher error and debugging output.

You can control logging on a per-channel basis or you can specify that message activity on all channels be logged. In the initial configuration, logging is disabled on all channels.

See Enabling MTA Logging for more information.

Enabling logging causes the MTA to write an entry to the msg_svr_base/data/log/mail* file each time a message passes through an MTA channel. Such log entries can be useful for gathering statistics on how many messages are passing through the MTA (or through particular channels). You can also use these log entries to investigate other issues, such as whether and when a message was sent or delivered.

The message return job, which runs every night around midnight, appends any existing mail.log_yesterday to the cumulative log file, mail.log, renames the current mail.log_current file to mail.log_yesterday, and then begins a new mail.log_current file. The message return job also performs the analogous operations for any connection.log* files.

While the MTA performs automatic rollovers to maintain the current file, you must manage the cumulative mail.log file by determining policies for tasks such as backing up the file, truncating the file, deleting the file, and so on.

When considering how to manage the log files, note that the MTA periodic return job will execute a site-supplied msg_svr_base/bin/daily_cleanup procedure, if one exists. Thus some sites might choose to supply their own cleanup procedure that, for instance, renames the old mail.log file once a week (or once a month), and so on.

With logging is enabled, the mail.log file steadily grows and, if left unchecked, consumes all available disk space. Monitor the size of this file and periodically delete unnecessary contents. You can also delete the entire file as another version will be created as needed.

Understanding the MTA Log Entry Format

The MTA log file is written as ASCII text. By default, each log file entry contains eight or nine fields as shown in the example below.

19-Jan-1998 19:16:57.64 l tcp_local E 1 adam@sesta.com rfc822;marlowe@siroe.com marlowe@siroe.com

The log entry shows:

1. The date and time the entry was made (in the example, 19-Jan-1998 19:16:57.64).

2. The channel name for the source channel (in the example, l).

3. The channel name for the destination channel (in the example, tcp_local). (For SMTP channels, when LOG_CONNECTION is enabled, a plus (+) indicates inbound to the SMTP server; a minus (-) indicates outbound via the SMTP client.)

4. The type of entry (in the example, E); see Table 21–2.

5. The size of the message (in the example, 1). This is expressed in kilobytes by default, although this default can be changed by using the BLOCK_SIZE keyword in the MTA option file.

6. The envelope From: address (in the example, adam@sesta.com). Note that for messages with an empty envelope From: address, such as notification messages, this field is blank.

7. The original form of the envelope To: address (in the example, marlowe@siroe.com).

8. The active (current) form of the envelope To: address (in the example, marlowe@siroe.com).

9. The delivery status (SMTP channels only).

The following table describes the logging entry codes.

With LOG_CONNECTION, LOG_FILENAME, LOG_MESSAGE_ID, LOG_NOTARY, LOG_PROCESS, and LOG_USERNAME all enabled in the MTA Option file, the format becomes as shown in the example below. (The sample log entry line has been wrapped for typographic reasons; the actual log entry would appear on one physical line.)

19-Jan-1998 13:13:27.10 HOSTA   2e2d.2.1 tcp_local   l
 E 1 service@siroe.com rfc822;adam@sesta.com
 adam 276 /imta/queue/l/ZZ01IWFY9ELGWM00094D.00
 <01IWFVYLGTS499EC9Y@siroe.com> inetmail
 siroe.com (siroe.com [192.160.253.66])
                  

Where the additional fields, beyond those already discussed above, are:

1. The name of the node on which the channel process is running (in the example, HOSTA).

2. The process ID (expressed in hexadecimal), followed by a period (dot) character and a count. If this had been a multithreaded channel entry (that is, a tcp_* channel entry), there would also be a thread ID present between the process ID and the count. In the example, the process ID is 2e2d.2.1.

3. The NOTARY (delivery receipt request) flags for the message, expressed as an integer (in the example, 276).

4. The file name in the MTA queue area (in the example, /imta/queue/l/ZZ01IWFY9ELGWM00094D.00).

5. The message ID (in the example, <01IWFVYLGTS499EC9Y@siroe.com>).

6. The name of the executing process (in the example, inetmail). On UNIX, for dispatcher processes such as the SMTP server, this will usually be inetmail (unless SASL was used).

7. The connection information (in the example, siroe.com (siroe.com [192.160.253.66]). The connection information consists of the sending system or channel name, such as the name presented by the sending system on the HELO/EHLO line (for incoming SMTP messages), or the enqueuing channel's official host name (for other sorts of channels). In the case of TCP/IP channels, the sending system's “real” name, that is, the symbolic name as reported by a DNS reverse lookup and/or the IP address, can also be reported within parentheses as controlled by the ident* channel keywords; see IDENT Lookups instance us of the default identnone keyword, that selects display of both the name found from the DNS and IP address.

Enabling MTA Logging

To gather statistics for just a few particular MTA channels, enable the logging channel keyword on just those MTA channels of interest. Many sites prefer to enable logging on all MTA channels. In particular, if you are trying to track down problems, the first step in diagnosing some problems is to notice that messages are not going to the channel you expected or intended, and having logging enabled for all channels can help you investigate such problems.

To Enable MTA Logging on a Specific Channel

Steps

1. Edit the imta.cnf file.

   The file is located in the /opt/SUNWmsgsr/config directory.

2. To enable logging for a particular channel, add the logging keyword to the channel definition. For example:

   channel-name keyword1 keyword2 logging

   In addition, you can also set a number of configuration parameters such as directory path for log files, log levels, an so on. See Managing Message Store, Admin, and Default Service Logs

To Enable MTA Logging on All Channels

Steps

1. Edit the imta.cnf file.

   The file is located in the /opt/SUNWmsgsr/config directory.

2. Add the logging keyword to your defaults channel (see Configuring Channel Defaults configuration file. For example:

   defaults logging notices 1 2 4 7 copywarnpost copysendpost postheadonly noswitchchannel immnonurgent maxjobs 7 defaulthost siroe.com l defragment charset7 us-ascii charset8 iso-8859-01 siroe.com

Specifying Additional MTA Logging Options

In addition to the basic information always provided when logging is enabled, you can specify that additional, optional information fields be included by setting various LOG_* MTA options in the MTA Option file. The file specified with the IMTA_OPTION_FILE option in the IMTA tailor file (msg_svr_base/config/imta_tailor) specifies the MTA Option file. By default, this is the msg_svr_base/config/option.dat file.

For complete details about the MTA Option file, see the Option File in Sun Java System Messaging Server 6 2005Q4 Administration Reference.

To Send MTA Logs to syslog

Steps

1. Edit the MTA Option file.

2. Set the LOG_MESSAGES_SYSLOG option to 1.

   A value of 0 is the default and indicates that syslog (event log) logging is not performed.

To Correlate Log Message Entries

Steps

1. Edit the MTA Option file.

2. Set the LOG_MESSAGE_ID option to 1.

   A value of 0 is the default and indicates that message IDs are not saved in the mail.log file.

To Identify Message Delivery Retries

Steps

1. Edit the MTA Option file.

2. Set the LOG_FILENAME option to 1.

   This option makes it easier to immediately spot how many times the delivery of a particular message file has been retried. This option can also be useful in understanding when the MTA does or does not split a message to multiple recipients into separate message file copies on disk.

To Log TCP/IP Connections

Steps

1. Edit the MTA Option file.

2. Set the LOG_CONNECTION option.

   This option causes the MTA to log TCP/IP connections, as well as message traffic. The connection log entries are written to the mail.log* files by default. Optionally, the connection log entries can be written to connection.log* files. See the SEPARATE_CONNECTION_LOG option for more information.

To Write Entries to the connection.log File

Steps

1. Edit the MTA Option file.

2. Set the SEPARATE_CONNECTION_LOG option to 1.

   Use this option to specify that connection log entries instead be written to connection.log files. The default value of 0 causes the connection logging to be stored in the MTA log files.

To Correlate Log Messages by Process ID

Steps

1. Edit the MTA Option file.

2. Set the LOG_PROCESS option.

   When used in conjunction with LOG_CONNECTION, this option enables correlation by process ID of which connection entries correspond to which message entries.

To Save User Names Associated with a Process That Enqueues Mail to the mail.log File

Steps

1. Edit the MTA Option file.

2. Set the LOG_USERNAME option.

   This option controls whether or not the user name associated with a process that enqueues mail is saved in the mail.log file. For SMTP submissions where SASL (SMTP AUTH) is used, the user name field will be the authenticated user name (prefixed with an asterisk character).

MTA Message Logging Examples

The exact field format and list of fields logged in the MTA message files vary according to the logging options set. This section shows a few examples of interpreting typical sorts of log entries. For a description of additional, optional fields, see Specifying Additional MTA Logging Options.

For typographic reasons, log file entries will be shown folded onto multiple lines—actual log file entries are one line per entry.

When reviewing a log file, keep in mind that on a typical system many messages are being handled at once. Typically, the entries relating to a particular message will be interspersed among entries relating to other messages being processed during that same time. The basic logging information is suitable for gathering a sense of the overall numbers of messages moving through the MTA.

If you wish to correlate particular entries relating to the same message to the same recipient(s), enable LOG_MESSAGE_ID. To correlate particular messages with particular files in the MTA queue area, or to see from the entries how many times a particular not-yet-successfully-dequeued message has had delivery attempted, enable LOG_FILENAME. For SMTP messages (handled via a TCP/IP channel), if you want to correlate TCP connections to and from remote systems with the messages sent, enable LOG_PROCESS and some level of LOG_CONNECTION.

MTA Logging Example: User Sends an Outgoing Message

The example below shows a fairly basic example of the sorts of log entries one might see if a local user sends a message out an outgoing TCP/IP channel, for example, to the Internet. In this example, LOG_CONNECTION is enabled. The lines marked with (1) and (2) are one entry—they would appear on one physical line in an actual log file. Similarly, the lines marked with (3) - (7) are one entry and would appear on one physical line.

Example 21–1 MTA Logging: A Local User Sends An Outgoing Message

19-Jan-1998 19:16:57.64 l            tcp_local    E 1         (1)
 adam@sesta.com rfc822;marlowe@siroe.com marlowe@siroe.com     (2)
 
 19-Jan-1998 19:17:01.16 tcp_local                  D 1        (3)
 adam@sesta.com rfc822;marlowe@siroe.com marlowe@siroe.com     (4)
 dns;thor.siroe.com
 (TCP|206.184.139.12|2788|192.160.253.66|25)                   (5)
 (THOR.SIROE.COM -- Server ESMTP [iMS V5.0 #8694])             (6)
 smtp;250 2.1.5 marlowe@siroe.com and options OK.              (7)

1. This line shows the date and time of an enqueue (E) from the l channel to the tcp_local channel of a one (1) block message.

2. This is part of the same physical line of the log file as (1), presented here as a separate line for typographical convenience. It shows the envelope From: address, in this case adam@sesta.com, and the original version and current version of the envelope To: address, in this case marlowe@siroe.com.

3. This shows the date and time of a dequeue (D) from the tcp_local channel of a one (1) block message—that is, a successful send by the tcp_local channel to some remote SMTP server.

4. This shows the envelope From: address, the original envelope To: address, and the current form of the envelope To: address.

5. This shows that the actual system to which the connection was made is named thor.siroe.com in the DNS, that the local sending system has IP address 206.184.139.12 and is sending from port 2788, that the remote destination system has IP address 192.160.253.66 and the connection port on the remote destination system is port 25.

6. This shows the SMTP banner line of the remote SMTP server.

7. This shows the SMTP status code returned for this address; 250 is the basic SMTP success code and in addition, this remote SMTP server responds with extended SMTP status codes and some additional text.

MTA Logging Example: Including Optional Logging Fields

This example shows a logging entry similar to that shown in Example 21–3 and LOG_MESSAGE_ID=1 showing the file name and message ID; see (1) and (2). The message ID in particular can be used to correlate which entries relate to which message.

Example 21–2 MTA Logging – Including Optional Logging Fields

19-Jan-1998 19:16:57.64 l            tcp_local    E 1
  adam@sesta.com rfc822;marlowe@siroe.com marlowe@siroe.com
  /imta/queue/tcp_local/ZZ01ISKLSKLZLI90N15M.00
   <01ISKLSKC2QC90N15M@sesta.com>    (1)
                  
 19-Jan-1998 19:17:01.16 tcp_local                  D 1
  adam@sesta.com rfc822;marlowe@siroe.com marlowe@siroe.com
  /imta/queue/tcp_local/Z01ISKLSKLZLI90N15M.00
    <01ISKLSKC2QC90N15M@sesta.com>   (2)
  dns;thor.siroe.com (TCP|206.184.139.12|2788|192.160.253.66|25)
  (THOR.SIROE.COM -- Server ESMTP [iMS V5.0 #8694])
  smtp;250 2.1.5 marlowe@siroe.com and options OK.

MTA Logging Example – Sending to a List

This example illustrates sending to multiple recipients with LOG_FILENAME=1, LOG_MESSAGE_ID=1, and LOG_CONNECTION=1 enabled. Here user adam@sesta.com has sent to the MTA mailing list test-list@sesta.com, which expanded to bob@sesta.com, carol@varrius.com, and david@varrius.com. Note that the original envelope To: address is test-list@sesta.com for each recipient, though the current envelope To: address is each respective address. Note how the message ID is the same throughout, though two separate files (one for the l channel and one going out the tcp_local channel) are involved.

Example 21–3 MTA Logging – Sending to a List

19-Jan-1998 20:01:44.10 l    l                    E 1
  adam@sesta.com rfc822;test-list@sesta.com bob
  imta/queue/l/ZZ01ISKND3DE1K90N15M.00
  <01ISKND2H8MS90N15M@sesta.com>
 
 19-Jan-1998 20:01:44.81 l            tcp_local    E 1
  adam@sesta.com rfc822;test-list@sesta.com carol@varrius.com
  imta/queue/tcp_local/ZZ01ISKND2WS1I90N15M.00 
  <01ISKND2H8MS90N15M@sesta.com>
 
 19-Jan-1998 20:01:44.81 l            tcp_local    E 1
  adam@sesta.com rfc822;test-list@sesta.com david@varrius.com
  imta/queue/tcp_local/ZZ01ISKND2WS1I90N15M.00
 <01ISKND2H8MS90N15M@sesta.com>
 
 19-Jan-1998 20:01:50.69 l                         D 1
  adam@sesta.com rfc822;test-list@sesta.com bob
  imta/queue/l/ZZ01ISKND3DE1K90N15M.00
  <01ISKND2H8MS90N15M@sesta.com>
 
 19-Jan-1998 20:01:57.36 tcp_local                  D 1
  adam@sesta.com rfc822;test-list@sesta.com carol@varrius.com
  imta/queue/tcp_local/ZZ01ISKND2WS1I90N15M.00
  <01ISKND2H8MS90N15M@sesta.com>
  dns;gw.varrius.com (TCP|206.184.139.12|2788|192.160.253.66|25)
  (gw.varrius.com -- SMTP Sendmail)
  smtp;250 OK.
 
 19-Jan-1998 20:02:06.14 tcp_local                  D 1
  adam@sesta.com rfc822;test-list@sesta.com david@varrius.com
  imta/queue/tcp_local/ZZ01ISKND2WS1I90N15M.00
  <01ISKND2H8MS90N15M@sesta.com>
  dns;gw.varrius.com (TCP|206.184.139.12|2788|192.160.253.66|25)
  (gw.varrius.com -- SMTP Sendmail)
  smtp;250 OK.

MTA Logging – Sending to a Nonexistent Domain

This example illustrates an attempt to send to a nonexistent domain (here very.bogus.com); that is, sending to a domain name that is not noticed as nonexistent by the MTA's rewrite rules and that the MTA matches to an outgoing TCP/IP channel. This example assumes the MTA option settings of LOG_FILENAME=1 and LOG_MESSAGE_ID=1.

When the TCP/IP channel runs and checks for the domain name in the DNS, the DNS returns an error that no such name exists. Note the “rejection” entry (R), as seen in (5), with the DNS returning an error that this is not a legal domain name, as seen in (6).

Because the address is rejected after the message has been submitted, the MTA generates a bounce message to the original sender. The MTA enqueues the new rejection message to the original sender (1), and sends a copy to the postmaster (4) before deleting the original outbound message (the R entry shown in (5)).

Notification messages, such as bounce messages, have an empty envelope From: address—as seen, for instance, in (2) and (8)—in which the envelope From: field is shown as an empty space. The initial enqueue of a bounce message generated by the MTA shows the message ID for the new notification message followed by the message ID for the original message (3). (Such information is not always available to the MTA, but when it is available to be logged, it allows correlation of the log entries corresponding to the outbound failed message with the log entries corresponding to the resulting notification message.) Such notification messages are enqueued to the process channel, which in turn enqueues them to an appropriate destination channel (7).

Example 21–4 MTA Logging – Sending to a Nonexistent Domain

19-JAN-1998 20:49:04 l            tcp_local    E 1
  adam@sesta.com rfc822;user@very.bogus.com user@very.bogus.com
  imta/queue/tcp_local/ZZ01ISKP0S0LVQ94DU0K.00
 <01ISKP0RYMAS94DU0K@SESTA.COM>
 
19-JAN-1998 20:49:33 tcp_local    process      E 1                (1)
 rfc822;adam@sesta.com adam@sesta.com                             (2)
 imta/queue/process/ZZ01ISKP0S0LVQ94DTZB.00
 <01ISKP22MW8894DTAS@SESTA.COM>,<01ISKP0RYMAS94DU0K@SESTA.COM>    (3)

19-JAN-1998 20:49:33 tcp_local    process      E 1                (4)
 rfc822;postmaster@sesta.com postmaster@sesta.com
 imta/queue/process/ZZ01ISKP0S0LVQ94DTZB.00
 <01ISKP22MW8894DTAS@SESTA.COM>,<01ISKP0RYMAS94DU0K@SESTA.COM>
 
19-JAN-1998 20:50:07 tcp_local                  R 1               (5)
 adam@sesta.com rfc822;user@very.bogus.com user@very.bogus.com
 imta/queue/tcp_local/ZZ01ISKP0S0LVQ94DU0K.00
 <01ISKP0RYMAS94DU0K@SESTA.COM>
 Illegal host/domain name found                                   (6)

19-JAN-1998 20:50:08 process      l            E 3                (7)
 rfc822;adam@sesta.com adam                                       (8)
 imta/queue/l/ZZ01ISKP23BUQS94DTYL.00
 <01ISKP22MW8894DTAS@SESTA.COM>
 
19-JAN-1998 20:50:08 process      l            E 3
  rfc822;postmaster@sesta.com postmaster
  imta/queue/l/ZZ01ISKP23BUQS94DTYL.00
 <01ISKP22MW8894DTAS@SESTA.COM>
 
19-JAN-1998 20:50:12 l                         D 3
  rfc822;adam@sesta.com adam
  imta/queue/l/ZZ01ISKP23BUQS94DTYL.00
  <01ISKP22MW8894DTAS@SESTA.COM>
 
19-JAN-1998 20:50:12 l                         D 3
  rfc822;postmaster@sesta.com postmaster
  imta/queue/l/ZZ01ISKP23BUQS94DTYL.00
  <01ISKP22MW8894DTAS@SIROE.COM>

MTA Logging Example – Sending to a Nonexistent Remote User

This example illustrates an attempt to send to a bad address on a remote system. This example assumes MTA option settings of LOG_FILENAME=1 and LOG_MESSAGE_ID=1, and channel option settings of LOG_BANNER=1 and LOG_TRANSPORTINFO=1. Note the rejection entry (R), seen in (1). But in contrast to the rejection entry in Example 21–4, note that the rejection entry here shows that a connection to a remote system was made, and shows the SMTP error code issued by the remote SMTP server, (2) and (3). The inclusion of the information shown in (2) is due to setting the channel options LOG_BANNER=1 and LOG_TRANSPORTINFO=1.

Example 21–5 MTA Logging – Sending to a Nonexistent Remote User

20-JAN-1998 13:11:05 l            tcp_local    E 1
  adam@sesta.com rfc822;nonesuch@siroe.com nonesuch@siroe.com
  imta/queue/tcp_local/ZZ01ISLNBB1JOE94DUWH.00
  <01ISLNBAWV3094DUWH@sesta.com>
 
20-JAN-1998 13:11:08 tcp_local    process      E 1
  rfc822;adam@sesta.com adam@sesta.com
  imta/queue/process/ZZ01ISLNBB1JOE94DSGB.00
  <01ISLNBFKIDS94DUJ8@sesta.com>,<01ISLNBAWV3094DUWH@sesta.com>
 
 20-JAN-1998 13:11:08 tcp_local    process      E 1
  rfc822;postmaster@sesta.com postmaster@sesta.com
  imta/queue/process/ZZ01ISLNBB1JOE94DSGB.00
  <01ISLNBFKIDS94DUJ8@sesta.com>,<01ISLNBAWV3094DUWH@sesta.com>
 
20-JAN-1998 13:11:11 tcp_local                  R 1       (1)
  adam@sesta.com rfc822;nonesuch@siroe.com nonesuch@siroe.com
  imta/queue/tcp_local/ZZ01ISLNBB1JOE94DUWH.00 
  <01ISLNBAWV3094DUWH@sesta.com>
  dns;thor.siroe.com
  (TCP|206.184.139.12|2788|192.160.253.66|25)             (2)
  (THOR.SIROE.COM -- Server ESMTP [iMS V5.0 #8694])
  smtp; 553 unknown or illegal user: nonesuch@siroe.com   (3)
 
20-JAN-1998 13:11:12 process      l            E 3
  rfc822;adam@sesta.com adam
  imta/queue/l/ZZ01ISLNBGND1094DQDP.00
  <01ISLNBFKIDS94DUJ8@sesta.com>
 
20-JAN-1998 13:11:12 process      l            E 3
  rfc822;postmaster@sesta.com postmaster
  imta/queue/l/ZZ01ISLNBGND1094DQDP.00
  <01ISLNBFKIDS94DUJ8@sesta.com>
 
20-JAN-1998 13:11:13 l                         D 3
  rfc822;adam@sesta.com adam@sesta.com
  imta/queue/l/ZZ01ISLNBGND1094DQDP.00
  <01ISLNBFKIDS94DUJ8@sesta.com>
 
20-JAN-1998 13:11:13 l                         D 3
  rfc822;postmaster@sesta.com postmaster@sesta.com
  imta/queue/l/ZZ01ISLNBGND1094DQDP.00
  <01ISLNBFKIDS94DUJ8@sesta.com>

MTA Logging Example – Rejecting a Remote Side's Attempt to Submit a Message

This example illustrates the sort of log file entry resulting when the MTA rejects a remote side’s attempt to submit a message. (This example assumes that no optional LOG_* options are enabled, so only the basic fields are logged in the entry. Note that enabling the LOG_CONNECTION option, in particular, would result in additional informative fields in such J entries.) In this case, the example is for an MTA that has set up SMTP relay blocking (see Configuring SMTP Relay Blocking) with an ORIG_SEND_ACCESS mapping, including:

ORIG_SEND_ACCESS

! ...numerous entries omitted...
!
   tcp_local|*|tcp_local|*   $NRelaying$ not$ permitted

and where alan@very.bogus.com is not an internal address. Hence the attempt of the remote user harold@varrius.com to relay through the MTA system to the remote user alan@very.bogus.com is rejected.

Example 21–6 MTA Logging – Rejecting a Remote Side's Attempt to Submit a Message

28-May-1998 12:02:23 tcp_local            J 0               (1)
 harold@varrius.com rfc822; alan@very.bogus.com             (2)
 550 5.7.1 Relaying not permitted: alan@very.bogus.com      (3)

1. This log shows the date and time the MTA rejects a remote side’s attempt to submit a message. The rejection is indicated by a J record. (Cases where an MTA channel is attempting to send a message which is rejected is indicated by R records, as shown in Example 21–4 and Example 21–5).

   ---

   Note –

   The last J record written to the log will have an indication stating that it is the last for a given session. Also, the current version of Messaging Server does not place a limit on the number of J records.

   ---

2. The attempted envelope From: and To: addresses are shown. In this case, no original envelope To: information was available so that field is empty.

3. The entry includes the SMTP error message the MTA issued to the remote (attempted sender) side.

MTA Logging Example – Multiple Delivery Attempts

This example illustrates the sort of log file entries resulting when a message cannot be delivered upon the first attempt, so the MTA attempts to send the message several times. This example assumes option settings of LOG_FILENAME=1 and LOG_MESSAGE_ID=1.

Example 21–7 MTA Logging – Multiple Delivery Attempts

15-Jan-1998 10:31:05.18 tcp_internal   tcp_local   E 3          (1)
 adam@hosta.sesta.com rfc822;user@some.org user@some.org
 imta/queue/tcp_local/ZZ01IS3D2ZP7FQ9UN54R.00 
 <01IRUD7SVA3Q9UN2D4@sesta.com>
 
15-Jan-1998 10:31:10.37 tcp_local                  Q 3          (2)
 adam@hosta.sesta.com rfc822;user@some.org user@some.org
 imta/queue/tcp_local/ZZ01IS3D2ZP7FQ9UN54R.00                   (3)
 <01IRUD7SVA3Q9UN2D4@sesta.com>
 TCP active open: Failed connect()    Error: no route to host   (4)
 
     ...several hours worth of entries...

15-Jan-1998 12:45:39.48 tcp_local                  Q 3          (5)
  adam@hosta.sesta.com rfc822;user@some.org user@some.org
  imta/queue/tcp_local/ZY01IS3D2ZP7FQ9UN54R.00                  (6)
  <01IRUD7SVA3Q9UN2D4@sesta.com>
  TCP active open: Failed connect()    Error: no route to host
 
  ...several hours worth of entries...

 15-Jan-1998 16:45:24.72 tcp_local                  Q 3
  adam@hosta.sesta.com rfc822;user@some.org user@some.org
  imta/queue/tcp_local/ZX01IS67NY4RRK9UN7GP.00                  (7)
                   <01IRUD7SVA3Q9UN2D4@sesta.com>
  TCP active open: Failed connect() Error: connection refused   (8)

  ...several hours worth of entries...

 15-Jan-1998 20:45:51.55 tcp_local                  D 3         (9)
  adam@hosta.sesta.com rfc822;user@some.org user@some.org
  imta/queue/tcp_local/ZX01IS67NY4RRK9UN7GP.00
  <01IRUD7SVA3Q9UN2D4@sesta.com>
  dns;host.some.org (TCP|206.184.139.12|2788|192.1.1.1|25)
  (All set, fire away)
  smtp; 250 Ok

1. The message comes in the tcp_internal channel—perhaps from a POP or IMAP client, or perhaps from another host within the organization using the MTA as an SMTP relay; the MTA enqueues it to the outgoing tcp_local channel.

2. The first delivery attempt fails, as indicated by the Q entry.

3. That this is a first delivery attempt can be seen from the ZZ* filename.

4. This delivery attempt failed when the TCP/IP package could not find a route to the remote side. As opposed to Example 21–4, the DNS did not object to the destination domain name, some.org; rather, the “no route to host” error indicates that there is some network problem between the sending and receiving side.

5. The next time the MTA periodic job runs it reattempts delivery, again unsuccessfully.

6. The file name is now ZY*, indicating that this is a second attempt.

7. The file name is ZX* for this third unsuccessful attempt.

8. The next time the periodic job reattempts delivery the delivery fails, though this time the TCP/IP package is not complaining that it cannot get through to the remote SMTP server, but rather the remote SMTP server is not accepting connections. (Perhaps the remote side fixed their network problem, but has not yet brought their SMTP server back up—or their SMTP server is swamped handling other messages and hence was not accepting connections at the moment the MTA tried to connect.)

9. Finally the message is dequeued.

MTA Logging – Incoming SMTP Message Routed Through the Conversion Channel

This example illustrates the case of a message routed through the conversion channel. The site is assumed to have a CONVERSIONS mapping table such as:

CONVERSIONS
  IN-CHAN=tcp_local;OUT-CHAN=l;CONVERT   Yes

This example assumes option settings of LOG_FILENAME=1 and LOG_MESSAGE_ID=1.

Example 21–8 MTA Logging – Incoming SMTP Message Routed Through the Conversion Channel

04-Feb-1998 00:06:26.72 tcp_local    conversion   E 9      (1)
  amy@siroe.edu rfc822;bert@sesta.com bert@sesta.com
  imta/queue/conversion/ZZ01IT5UAMZ4QW98518O.00
  <01IT5UALL14498518O@siroe.edu>
 
 04-Feb-1998 00:06:29.06 conversion   l            E 9     (2)
                   amy@siroe.edu rfc822;bert@sesta.com bert
  imta/queue/l/ZZ01IT5UAOXLDW98509E.00  <01IT5STUMUFO984Z8L@siroe.edu>
 
04-Feb-1998 00:06:29.31 conversion                D 9      (3)
 amy@siroe.edu rfc822;bert@sesta.com bert
 imta/queue/conversion/ZZ01IT5UAMZ4QW98518O.00
 <01IT5UALL14498518O@siroe.edu>
 
 04-Feb-1998 00:06:32.62 l                         D 9     (4)
  amy@siroe.edu rfc822;bert@siroe.com bert
  imta/queue/l/ZZ01IT5UAOXLDW98509E.00
  <01IT5STUMUFO984Z8L@siroe.edu>

               

1. The message from external user amy@siroe.edu comes in addressed to the l channel recipient bert@sesta.com. The CONVERSIONS mapping entry, however, causes the message to be initially enqueued to the conversion channel (rather than directly to the l channel).

2. The conversion channel runs and enqueues the message to the l channel.

3. Then the conversion channel can dequeue the message (delete the old message file).

4. And finally the l channel dequeues (delivers) the message.

MTA Logging Example: Outbound Connection Logging

This example illustrates log output for an outgoing message when connection logging is enabled, via LOG_CONNECTION=3. LOG_PROCESS=1, LOG_MESSAGE_ID=1 and LOG_FILENAME=1 are also assumed in this example. The example shows the case of user adam@sesta.com sending the same message (note that the message ID is the same for each message copy) to three recipients, bobby@hosta.sesta.com, carl@hosta.sesta.com, and dave@hostb.sesta.com. This example assumes that the message is going out a tcp_local channel marked (as such channels usually are) with the single_sys channel keyword. Therefore, a separate message file on disk will be created for each set of recipients to a separate host name, as seen in (1), (2), and (3), where the bobby@hosta.sesta.com and carl@hosta.sesta.com recipients are stored in the same message file, but the dave@hostb.sesta.com recipient is stored in a different message file.

Example 21–9 MTA Logging: Outbound Connection Logging

19-Feb-1998 10:52:05.41 1e488.0 l            tcp_local    E 1
 adam@sesta.com rfc822;bobby@hosta.sesta.com bobby@hosta.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7BO388000FCN.00                    (1)
  <01ITRF7BDHS6000FCN@SESTA.COM>

19-Feb-1998 10:52:05.41 1e488.0 l            tcp_local    E 1
 adam@sesta.com rfc822;carl@hosta.sesta.com carl@hosta.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7BO388000FCN.00                    (2)
   <01ITRF7BDHS6000FCN@SESTA.COM>

19-Feb-1998 10:52:05.74 1e488.1 l            tcp_local    E 1
 adam@sesta.com rfc822;dave@hostb.sesta.com dave@hostb.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7C11FU000FCN.00                    (3)
   <01ITRF7BDHS6000FCN@SESTA.COM>

19-Feb-1998 10:52:10.79 1f625.2.0 tcp_local    -            O    (4)
 TCP|206.184.139.12|5900|206.184.139.66|25
 SMTP/hostb.sesta.com/mailhub.sesta.com                          (5)
 
19-Feb-1998 10:52:10.87 1f625.3.0 tcp_local    -            O    (6)
 TCP|206.184.139.12|5901|206.184.139.70|25
 SMTP/hosta.sesta.com/hosta.sesta.com                            (7)

19-Feb-1998 10:52:12.28 1f625.3.1 tcp_local                  D 1
 adam@sesta.com rfc822;bobby@hosta.sesta.com bobby@hosta.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7BO388000FCN.00
  <01ITRF7BDHS6000FCN@SESTA.COM>
 hosta.sesta.com dns;hosta.sesta.com                            (8)
 (TCP|206.184.139.12|5901|206.184.139.70|25)
 (hosta.sesta.com -- Server ESMTP [iMS V5.0 #8790])
 (TCP|206.184.139.12|5901|206.184.139.70|25)
 smtp;250 2.1.5 bobby@hosta.sesta.com and options OK.

19-Feb-1998 10:52:12.28 1f625.3.1 tcp_local                  D 1
 adam@sesta.com rfc822;carl@hosta.sesta.com carl@hosta.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7BO388000FCN.00
  <01ITRF7BDHS6000FCN@SESTA.COM>
 hosta.sesta.com dns;hosta.sesta.com
 (TCP|206.184.139.12|5901|206.184.139.70|25)
 (hosta.sesta.com -- Server ESMTP [iMS V5.0 #8790])
 (TCP|206.184.139.12|5901|206.184.139.70|25)
 smtp;250 2.1.5 carl@hosta.sesta.com and options OK.

19-Feb-1998 10:52:12.40 1f625.3.2 tcp_local      -            C  (9)
 TCP|206.184.139.12|5901|206.184.139.70|25
 SMTP/hosta.sesta.com/hosta.sesta.com

19-Feb-1998 10:52:13.01 1f625.2.1 tcp_local                  D 1
 adam@sesta.com rfc822;dave@hostb.sesta.com dave@hostb.sesta.com
 imta/queue/tcp_local/ZZ01ITRF7C11FU000FCN.00
  <01ITRF7BDHS6000FCN@SESTA.COM>
 mailhub.sesta.com dns;mailhub.sesta.com
 (TCP|206.184.139.12|5900|206.184.139.66|25)
 (MAILHUB.SESTA.COM -- Server ESMTP [iMS V5.0 #8694])
 (TCP|206.184.139.12|5900|206.184.139.66|25)
 smtp;250 2.1.5 dave@hostb.sesta.com and options OK.

19-Feb-1998 10:52:13.05 1f625.2.2 tcp_local      -            C  (10)
                  
 TCP|206.184.139.12|5900|206.184.139.66|25
 SMTP/hostb.sesta.com/mailhub.sesta.com

               

1. The message is enqueued to the first recipient...

2. ....and to the second recipient...

3. ....and to the third recipient.

4. Having LOG_CONNECTION=3 set causes the MTA to write this entry. The minus, -, indicates that this entry refers to an outgoing connection. The O means that this entry corresponds to the opening of the connection. Also note that the process ID here is the same, 1f625, since the same process is used for the multithreaded TCP/IP channel for these separate connection opens, though this open is being performed by thread 2 vs. thread 3.

5. As there are two separate remote systems to which to connect, the multithreaded SMTP client in separate threads opens up a connection to each—the first in this entry, and the second shown in 7. This part of the entry shows the sending and destination IP numbers and port numbers, and shows both the initial host name, and the host name found by doing a DNS lookup. In the SMTP/initial-host/dns-host clauses, note the display of both the initial host name, and that used after performing a DNS MX record lookup on the initial host name: mailhub.sesta.com is apparently an MX server for hostb.sesta.com.

6. The multithreaded SMTP client opens up a connection to the second system in a separate thread (though the same process).

7. As there are two separate remote systems to which to connect, the multithreaded SMTP client in separate threads opens up a connection to each—the second in this entry, and the first shown above in 5. This part of the entry shows the sending and destination IP numbers and port numbers, and shows both the initial host name, and the host name found by doing a DNS lookup. In this example, the system hosta.sesta.com apparently receives mail directly itself.

8. Besides resulting in specific connection entries, LOG_CONNECTION=3 also causes inclusion of connection related information in the regular message entries, as seen here for instance.

9. Having LOG_CONNECTION=3 causes the MTA to write this entry. After any messages are dequeued, (the bobby and carl messages in this example), the connection is closed, as indicated by the C in this entry.

MTA Logging Example: Inbound Connection Logging

This example illustrates log output for an incoming SMTP message when connection logging is enabled, via LOG_CONNECTION=3.

Example 21–10 MTA Logging – Inbound Connection Logging

19-Feb-1998 17:02:08.70 tcp_local    +            O          (1)
 TCP|206.184.139.12|25|192.160.253.66|1244 SMTP              (2)

19-Feb-1998 17:02:26.65 tcp_local    l             E 1
 service@siroe.com rfc822;adam@sesta.com adam
 THOR.SIROE.COM (THOR.SIROE.COM [192.160.253.66])            (3)

 19-Feb-1998 17:02:27.05 tcp_local    +             C        (4)
                   TCP|206.184.139.12|25|192.160.253.66|1244 SMTP
 
19-Feb-1998 17:02:31.73 l                          D 1
  service@siroe.com rfc822;adam@sesta.com adam

1. The remote system opens a connection. The O character indicates that this entry regards the opening of a connection; the + character indicates that this entry regards an incoming connection.

2. The IP numbers and ports for the connection are shown. In this entry, the receiving system (the system making the log file entry) has IP address 206.184.139.12 and the connection is being made to port 25; the sending system has IP address 192.160.253.66 and is sending from port 1244.

3. In the entry for the enqueue of the message from the incoming TCP/IP channel (tcp_local) to the l channel recipient, note that information beyond the default is included since LOG_CONNECTION=3 is enabled. Specifically, the name that the sending system claimed on its HELO or EHLO line, the sending system’s name as found by a DNS reverse lookup on the connection IP number, and the sending system’s IP address are all logged; see Chapter 12, Configuring Channel Definitions behavior.

4. The inbound connection is closed. The C character indicates that this entry regards the closing of a connection; the + character indicates that this entry regards an incoming connection.

Enabling Dispatcher Debugging

Dispatcher error and debugging output (if enabled) are written to the file dispatcher.log in the MTA log directory. The dispatcher configuration information is specified in the msg_svr_base/imta/dispatcher.cnf file. A default configuration file is created at installation time and can be used without any changes made. However, if you want to modify the default configuration file for security or performance reasons, you can do so by editing the dispatcher.cnf file.

To Enable Dispatcher Error Debugging Output

Steps

1. Edit the dispatcher.cnf file.

2. Set the DEBUG option to -1.

   You can also set the logical or environmental variable IMTA_DISPATCHER_DEBUG (UNIX), which defines a 32-bit debug mask in hexadecimal, to the value FFFFFFFF. The table above describes the meaning of each bit.

To Set Dispatcher Parameters (Solaris)

The dispatcher services offered in the dispatcher configuration file affects requirements for various system parameters. The system’s heap size (datasize) must be enough to accommodate the dispatcher’s thread stack usage.

Steps

1. To display the heap size (that is, default datasize), use one of the following:

   The csh command:

   # limit

   The ksh command

   # ulimit -a

   The Solaris utility

   # sysdef

2. For each dispatcher service compute STACKSIZE*MAX_CONNS, and then add up the values computed for each service. The system’s heap size needs to be at least twice this number.

Managing Message Store, Admin, and Default Service Logs

This section describes logging for the Message Store (POP, IMAP, and HTTP), Admin, and Default services. (see Table 21–1).

For these services, you can use Console to specify log settings and to view logs. The settings you specify affect which and how many events are logged. You can use those settings and other characteristics to refine searches for logged events when you are analyzing log files.

This section contains the following subsections:

• Understanding Service Log Characteristics

• Understanding Service Log File Format

• Defining and Setting Service Logging Options

• Searching and Viewing Service Logs

• Using Message Tracing for Message Store Logging

• Message Store Logging Examples

Understanding Service Log Characteristics

This section describes the following log characteristics for the message store and administration services: logging levels, categories of logged events, filename conventions for logs, and log-file directories.

Logging Levels

The level, or priority, of logging defines how detailed, or verbose, the logging activity is to be. A higher priority level means less detail; it means that only events of high priority (high severity) are logged. A lower level means greater detail; it means that more events are recorded in the log file.

You can set the logging level separately for each service—POP, IMAP, HTTP, Admin, and Default by setting the logfile.service.loglevel configuration parameter (see Defining and Setting Service Logging Options). You can also use logging levels to filter searches for log events. Table 21–4 describes the available levels. These logging levels are a subset of those defined by the UNIX syslog facility.

When you select a particular logging level, events corresponding to that level and to all higher (less verbose) levels are logged. The default level of logging is Notice.

The more verbose the logging you specify, the more disk space your log files will occupy; for guidelines, see Defining and Setting Service Logging Options

Categories of Logged Events

Within each supported service or protocol, Messaging Server further categorizes logged events by the facility, or functional area, in which they occur. Every logged event contains the name of the facility that generated it. These categories aid in filtering events during searches. Table 21–5 lists the categories that Messaging Server recognizes for logging purposes.

For examples of using categories as filters in log searches, see Searching and Viewing Service Logs

Service Log File Directories

Every logged service is assigned a single directory, in which its log files are stored. All IMAP log files are stored together, as are all POP log files, and log files of any other service. You define the location of each directory, and you also define how many log files of what maximum size are permitted to exist in the directory.

Make sure that your storage capacity is sufficient for all your log files. Log data can be voluminous, especially at lower (more verbose) logging levels.

It is important also to define your logging level, log rotation, log expiration, and server-backup policies appropriately so that all of your log-file directories are backed up and none of them become overloaded; otherwise, you may lose information. See Defining and Setting Service Logging Options.

Understanding Service Log File Format

All message store and administration service log files created by Messaging Server have identical content formats. Log files are multiline text files, in which each line describes one logged event. All event descriptions, for each of the supported services, have the general format:

dateTime hostName processName[pid]: category logLevel: eventMessage

Table 21–6 lists the log file components. Note that this format of event descriptions is identical to that defined by the UNIX syslog facility, except that the date/time format is different and the format includes two additional components (category and logLevel).

Here are three examples of logged events as viewed using Console:

02/May/1998:17:37:32 -0700 showshoe cgi_store[18753]:
 General Notice:
   Log created (894155852)

04/May/1998:11:07:44 -0400 xyzmail cgi_service[343]: General Error:
   function=getserverhello|port=2500|error=failed to connect

03/Dec/1998:06:54:32 +0200 SiroePost imapd[232]: Account Notice:
   close [127.0.0.1] [unauthenticated] 1998/12/3 6:54:32
   0:00:00 0 115 0

IMAP and POP event entries may end with three numbers. The example above has 0 115 0. The first number is bytes sent by client, the second number is the bytes sent by the server, and third number is mailboxes selected (always 1 for POP).

When viewing a log file in the Log Viewer window, you can limit the events displayed by searching for any specific component in an event, such as a specific logging level or category, or a specific process ID. For more information, see Searching and Viewing Service Logs.

The event message of each log entry is in a format specific to the type of event being logged, that is, each service defines what content appears in any of its event messages. Many event messages are simple and self-evident; others are more complex.

Defining and Setting Service Logging Options

You can define the message store and administration service logging configurations that best serve your administration needs. This section discusses issues that may help you decide on the best configurations and policies, and it explains how to implement them.

Flexible Logging Architecture

The naming scheme for log files (service.sequenceNum.timeStamp) helps you to design a flexible log-rotation and backup policy. The fact that events for different services are written to different files makes it easier for you to isolate problems quickly. Also, because the sequence number in a filename is ever-increasing and the timestamp is always unique, later log files do not simply overwrite earlier ones after a limited set of sequence numbers is exhausted. Instead, older log files are overwritten or deleted only when the more flexible limits of age, number of files, or total storage are reached.

Messaging Server supports automatic rotation of log files, which simplifies administration and facilitates backups. You are not required to manually retire the current log file and create a new one to hold subsequent logged events. You can back up all but the current log file in a directory at any time, without stopping the server or manually notifying the server to start a new log file.

In setting up your logging policies, you can set options (for each service) that control limits on total log storage, maximum number of log files, individual file size, maximum file age, and rate of log-file rotation.

Planning the Options You Want

Keep in mind that you must set several limits, more than one of which might cause the rotation or deletion of a log file. Whichever limit is reached first is the controlling one. For example, if your maximum log-file size is 3.5 MB, and you specify that a new log be created every day, you may actually get log files created faster than one per day if log data builds up faster than 3.5 MB every 24 hours. Then, if your maximum number of log files is 10 and your maximum age is 8 days, you may never reach the age limit on log files because the faster log rotation may mean that 10 files will have been created in less than 8 days.

The following default values, provided for Messaging Server administration logs, may be a reasonable starting point for planning:

Maximum number of log files in a directory: 10

Maximum log-file size: 2 MB

Total maximum size permitted for all log files: 20 MB

Minimum free disk space permitted: 5 MB

Log rollover time: 1 day

Maximum age before expiration: 7 days

Level of logging: Notice

You can see that this configuration assumes that server-administration log data is predicted to accumulate at about 2 MB per day, backups are weekly, and the total space allotted for storage of admin logs is at least 25 MB. (These settings may be insufficient if the logging level is more verbose.)

For POP, IMAP or HTTP logs, the same values might be a reasonable start. If all services have approximately the same log-storage requirements as the defaults shown here, you might expect to initially plan for about 150 MB of total log-storage capacity. (Note that this is meant only as a general indication of storage requirements; your actual requirements may be significantly different.)

Understanding Logging Options

You can set options that control the message store logging configuration by using Console or the command line.

The optimal settings for these options depend on the rate at which log data accumulates. It may take between 4,000 and 10,000 log entries to occupy 1 MB of storage. At the more verbose levels of logging (such as Notice), a moderately busy server may generate hundreds of megabytes of log data per week. Here is one approach you can follow:

• Set a level of logging that is consistent with your storage limits—that is, a level that you estimate will cause log-data accumulation at approximately the rate you used to estimate the storage limit.

• Define the log file size so that searching performance is not impacted. Also, coordinate it with your rotation schedule and your total storage limit. Given the rate at which log entries accumulate, you might set a maximum that is slightly larger than what you expect to accumulate by the time a rotation automatically occurs. And your maximum file size times your maximum number of files might be roughly equivalent to your total storage limit.

  For example, if your IMAP log rotation is daily, your expected accumulation of IMAP log data is 3 MB per day, and your total storage limit for IMAP logs is 25 MB, you might set a maximum IMAP log-file size of 3.5 MB. (In this example, you could still lose some log data if it accumulated so rapidly that all log files hit maximum size and the maximum number of log files were reached.)

• If server backups are weekly and you rotate IMAP log files daily, you might specify a maximum number of IMAP log files of about 10 (to account for faster rotation if the individual log-size limit is exceeded), and a maximum age of 7 or 8 days.

• Pick a total storage limit that is within your hardware capacity and that coordinates with the backup schedule you have planned for the server. Estimate the rate at which you anticipate that log data will accumulate, add a factor of safety, and define your total storage limit so that it is not exceeded over the period between server backups.

  For example, if you expect to accumulate an average of 3 MB of IMAP log-file data per day, and server backups are weekly, you might specify on the order of 25 - 30 MB as the storage limit for IMAP logs (assuming that your disk storage capacity is sufficient).

• For safety, pick a minimum amount of free disk space that you will permit on the volume that holds the log files. That way, if factors other than log-file size cause the volume to fill up, old log files will be deleted before a failure occurs from attempting to write log data to a full disk.

Searching and Viewing Service Logs

Console provides a basic interface for viewing message store and administration log data. It allows for selecting individual log files and for performing flexible filtered searches of log entries within those files.

For a given service, log files are listed in chronological order. Once you have chosen a log file to search, you can narrow the search for individual events by specifying search parameters.

Search Parameters

These are the search parameters you can specify for viewing log data:

• A time period. You can specify the beginning and end of a specific time period to retrieve events from, or you can specify a number of days (before the present) to search. You might typically specify a range to look at logged events leading up to a server crash or other occurrence whose time you know of. Alternatively, you might specify a day range to look at only today’s events in the current log file.

• A level of logging. You can specify the logging level (see Logging Levels example, Critical to see why the server went down, or Error to locate failed protocol calls.

• A facility. You can specify the facility (see Categories of Logged Events that contains the problem; for example, Store if you believe a server crash involved a disk error, or Protocol if the problem lies in an IMAP protocol command error.

• A text search pattern. You can provide a text search pattern to further narrow the search. You can include any component of the event (see Understanding Service Log File Format search, such as event time, process name, process ID, and any part of the event message (such as remote host name, function name, error number, and so on) that you know defines the event or events you want to retrieve.

  Your search pattern can include the following special and wildcard characters:

* Any set of characters (example: *.com)

? Any single character (example: 199?)

[nnn] Any character in the set nnn (example: [aeiou])

[^nnn] Any character not in the set nnn (example: [^aeiou])

[n-m] Any character in the range n-m (example: [A-Z])

[^n-m] Any character not in the range n-m (example: [^0-9])

\ Escape character: place before *, ?, [, or ] to use them as literals

Note: Searches are case-sensitive.

Examples of combining logging level and facility in viewing logs might include the following:

• Specifying Account facility (and Notice level) to display failed logins, which may be useful when investigating potential security breaches

• Specifying Network facility (and all logging levels) to investigate connection problems

• Specifying all facilities (and Critical logging level) to look for basic problems in the functioning of the server

Working With Service Logs

This section describes how to work with service logs by using the configutil command, and Console for searching and viewing logs.

To Send Service Logs to syslog

Step

Run the configutil command with the syslogfacility option:

configutil -o logfile.service.syslogfacility -v value

where service is admin, pop, imap, imta, or http and value is user, mail, daemon, local0 to local7, or none.

Once the value is set, messages are logged to the syslog facility corresponding to the set value and all the other log file service options are ignored. When the option is not set or the value is none, logging uses the Messaging Server log files.

To Set Logging Options Using Console

Steps

1. Open the Messaging Server whose log file options you want to set.

2. Click the Configuration tab, open the Log Files folder in the left pane, and select the log files of a service (such as IMAP, HTTP, or Admin).

3. From the “Levels of detail” drop-down list, choose a logging level.

4. In the “Directory path for log files” field, enter the name of the directory to hold your log files.

5. In the “File size for each log” field, enter your maximum log-file size.

6. In the “Create new log every” field, enter a number for the log-rotation schedule.

7. In the “Number of logs per directory” and the “When a log is older than” fields, enter the maximum number of log files and a maximum age to coordinate with your backup schedule.

8. In the “When total log size exceeds” field, enter the total storage limit you want.

9. In the “When free disk space is less than” field, enter the minimum amount of free disk space you want to reserve.

To Disable HTTP Logging

If your system does not support HTTP message access, that is, Webmail, you can disable HTTP logging by setting the following variables. Do not set these variables if your system requires Webmail support (for example, Messenger Express).

• Run the following configutil commands:

  configutil -o service.http.enable -v no
  configutil -o service.http.enablesslport -v no

To Set the Server Log Level

Step

Run the following configutil command:

configutil -o logfile.service.loglevel -v level

where service is admin, pop, imap, imta, or http and loglevel is Nolog, Critical, Error, Warning, Notice, Information, or Debug.

To Specify a Directory Path for Server Log Files

Step

Run the following configutil command:

configutil -o logfile.service.logdir -v dirpath

To Specify a Maximum File Size for Each Service Log

Step

Run the following configutil command:

configutil -o logfile.service.maxlogfilesize -v size

where size specifies a number of bytes.

To Specify a Service Log Rotation Schedule

Step

Run the following configutil command:

configutil -o logfile.service.rollovertime -v number

where number specifies a number of seconds.

To Specify a Maximum Number of Service Log Files Per Directory

Step

Run the following configutil command:

configutil -o logfile.service.maxlogfiles -v number

where number specifies the maximum number of log files.

To Specify a Storage Limit

Step

Run the following configutil command:

configutil -o logfile.service.maxlogsize -v number

where number specifies a number in bytes.

To Specify the Minimum Amount of Free Disk Space to Reserve

Step

Run the following configutil command:

configutil -o logfile.service.minfreediskspace -v number

where number specifies a number in bytes.

To Specify an Age for Logs at Which They Expire

configutil -o logfile.service.expirytime -v number

where number specifies a number in seconds.

To Specify a Search and View Results

Follow these steps to search for logged events with specific characteristics belonging to a given service:

Steps

1. In Console, open the Messaging Server whose log files you want to inspect.

2. Follow either of these steps to display the Log Files Content tab for a given logged service:

   • Click the Tasks tab, then click “View service logs”, where service is the name of the logged service (such as “IMAP service” or “administration”).

   • Click the Configuration tab, then open the Log Files folder in the left pane and select the log files of a service (such as IMAP or Admin). Then click the Content tab in the right pane.

3. The Content tab for that logged service is displayed.

4. In the Log filename field, select the log file you want to examine.

5. Click the View selected log button to open the Log Viewer window.

6. In the Log Viewer window, specify your desired search parameters (described in the previous section, Search Parameters).

7. Click Update to perform the search and display the results in the Log entry field.

Using Message Tracing for Message Store Logging

You can use Message Store logging to trace messages by message ID in a way similar to how the MTA traces messages. Tracing messages in this fashion enables you to track the critical events of a message’s life cycle.

To trace messages in the Message Store logs, you need to configure message tracing in addition to the normal logging configuration. By default, message tracing is not enabled.

Message tracing will fill up a large amount of disk space. Do not enable this feature unless you have adequate disk space.

Message Store logging can track the following operations:

• append - The primary way the Message Store library adds a message to a folder. Tracing append shows messages entering the Message Store.

• fetch - The IMAP command that retrieves a message or part of a message for the end user. For message tracing, its meaning is extended to when any service retrieves a message for the end user to read.

  In message tracing, you might sometimes like to avoid tracking when the header of a message is read, so a body fetch is in reference to when some part of the body of a message is retrieved.

• Expunge: An IMAP term that is extended in this case to reference when any service removes a message from a user’s folder.

To Enable Message Tracing

Step

Run the following configutil command:

configutil -o local.msgrace.active -v “yes”

Message trace information is written to the default log for each process. The IMAP fetches appear in the imap log file. The ims_master appends appear in the ims_master channel log file.

To Redirect Message Tracing to a Single Log File

Step

To redirect message trace logging to a single “msgtrace” log file, you must use the configutil command to configure the log file parameters. The msgtrace log file, unlike the other log files, is configured locally. For example:

configutil -o "local.logfile.msgtrace.buffersize" -v "0" configutil -o "local.logfile.msgtrace.expirytime" -v "604800" configutil -o "local.logfile.msgtrace.flushinterval" -v "60" configutil -o "local.logfile.msgtrace.logdir" -v "/opt/SUNWmsgsr/data/log" configutil -o "local.logfile.msgtrace.loglevel" -v "Information" configutil -o "local.logfile.msgtrace.logtype" -v "NscpLog" configutil -o "local.logfile.msgtrace.maxlogfiles" -v "10" configutil -o "local.logfile.msgtrace.maxlogfilesize" -v "2097152" configutil -o "local.logfile.msgtrace.maxlogsize" -v "20971520" configutil -o "local.logfile.msgtrace.minfreediskspace" -v "5242880" configutil -o "local.logfile.msgtrace.rollovertime" -v "86400"

To Unconfigure Message Trace Logging

Step

To unconfigure the msgtrace log file, use the configutil command to remove all references to its configuration. For example:

configutil -o "local.logfile.msgtrace.buffersize" -v "" configutil -o "local.logfile.msgtrace.expirytime" -v "" configutil -o "local.logfile.msgtrace.flushinterval" -v "" configutil -o "local.logfile.msgtrace.logdir" -v "" configutil -o "local.logfile.msgtrace.loglevel" -v "" configutil -o "local.logfile.msgtrace.logtype" -v "" configutil -o "local.logfile.msgtrace.maxlogfiles" -v "" configutil -o "local.logfile.msgtrace.maxlogfilesize" -v "" configutil -o "local.logfile.msgtrace.maxlogsize" -v "" configutil -o "local.logfile.msgtrace.minfreediskspace" -v "" configutil -o "local.logfile.msgtrace.rollovertime" -v ""

To Configure LMTP Logging

Step

If you are using LMTP, and not using a single “msgtrace” log file, then you must also locally configure the tcp_lmtp_server log file. If you are not using LMTP, or are not using message trace, or are using message trace in the “msgtrace” log file, you do not need to initialize the LMTP Message Store side log. (LMTP already logs MTA information separately.) For example:

configutil -o "local.logfile.tcp_lmtp_server.buffersize" -v "0" configutil -o "local.logfile.tcp_lmtp_server.expirytime" -v "604800" configutil -o "local.logfile.tcp_lmtp_server.flushinterval" -v "60" configutil -o "local.logfile.tcp_lmtp_server.logdir" -v \ "/opt/SUNWmsgsr/data/log" configutil -o "local.logfile.tcp_lmtp_server.loglevel" -v "Information" configutil -o "local.logfile.tcp_lmtp_server.logtype" -v "NscpLog" configutil -o "local.logfile.tcp_lmtp_server.maxlogfiles" -v "10" configutil -o "local.logfile.tcp_lmtp_server.maxlogfilesize" -v "2097152" configutil -o "local.logfile.tcp_lmtp_server.maxlogsize" -v "20971520" configutil -o "local.logfile.tcp_lmtp_server.minfreediskspace" \ -v "5242880" configutil -o "local.logfile.tcp_lmtp_server.rollovertime" -v "86400"

Other Message Store Logging Features

Messaging Server provides a feature called telemetry that can capture a user’s entire IMAP or POP session into a file. This feature is useful for debugging client problems. For example, if a user complains that their message access client is not working as expected, this feature can be used to trace the interaction between the access client and Messaging Server. See Check User IMAP/POP Session by Using Telemetry

Message Store Logging Examples

The exact field format and list of fields logged in the Message Store log files vary according to the logging options set. This section shows a few examples of interpreting typical sorts of log entries.

Message Store Logging Example: Bad Password

When a user types an invalid password, “authentication” failure is logged, as opposed to a “user not found” message. The message “user not found” is the text passed to the client for security reasons, but the real reason (invalid password) is logged.

Example 21–11 Message Store Logging – Invalid Password

 [30/Aug/2004:16:53:05 -0700] vadar imapd[13027]: Account Notice: badlogin:
[192.18.126.64:40718] plaintext user1 authentication failure

Message Store Logging – Account Disabled

The following example shows why a user cannot log in due to a disabled account. Furthermore, the disabled account is clarified as “(inactive)” or “(hold).”

Example 21–12 Message Store Logging – Account Disabled

[30/Aug/2004:16:53:31 -0700] vadar imapd[13027]: Account Notice: badlogin: 
[192.18.126.64:40720] plaintext user3 account disabled (hold)

Message Store Logging Example: Message Appended

The following example shows an append message, which occurs when whenever a message is appended to a folder. The Message Store log records all messages entering the Message Store through the ims_master and lmtp channels. Records the “append” of user ID, folder, message size, and message ID.

Example 21–13 Message Store Logging – Append

[31/Aug/2004:16:33:14 -0700] vadar ims_master[13822]: Store Information:append:
user1:user/user1:659:<Roam.SIMC.2.0.6.1093995286.11265.user1@vadar.siroe.com>

Message Store Logging Example: Message Retrieved by a Client

The Message Store log writes a “fetch” message when a client retrieves a message. The Message Store log records all client fetches of at least one body part. Records the “fetch” of user ID, folder, and message-ID.

Example 21–14 Message Store Logging – Message Retrieved by a Client

[31/Aug/2004:15:55:26 -0700] vadar imapd[13729]: Store Information:
fetch:user1:user/user1:<Roam.SIMC.2.0.6.1093051161.3655.user1@vad.siroe.com>

Message Store Logging Example: Message Removed from a Folder

Example 21–15 Message Store Logging Example: Message Removed from a Folder

The Message Store writes an “expunge” message when an IMAP or POP message is removed from a folder (but not removed from the system). It is logged whether it is expunged by the user or a utility. Records the “expunge” of folder and message ID.

31/Aug/2004:16:57:36 -0700] vadar imexpire[13923]: Store Information:
expunge:user/user1:<Roam.SIMC.2.0.6.1090458838.2929.user1@vadar.siroe.com>

Message Store Logging Example: Duplicate Login Messages

If you configure message trace for one msgtrace log file, the normal “login” messages, which appear in the imap and pop log files, are duplicated in the msgtrace file. The following is a normal login message:

Example 21–16 Message Store Logging – Login

[30/Aug/2004:16:53:13 -0700] vadar imapd[13027]: Account Information: login
[192.18.126.64:40718] user1 plaintext