Action definition file for event log trapping (Windows only)

Format

server event-server-name retry-times retry-count retry-interval retry-interval trap-interval monitoring-interval#1 matching-level [0 | 1] filter-check-level [0 | 1] jp1event-send [0 | 1]#1 ext-attr-option extended-attribute-name#2 unicode-trap [0 | 1]#2 # filter filter log-type condition-statement-1 condition-statement-2 : condition-statement-n end-filter

To Page Top

File name

ntevent.conf

To Page Top

Storage destination directory

installation-folder\conf\event\

To Page Top

Description

Specifies the conditions for converting event log data into JP1 event and the event-log monitoring interval.

To Page Top

Application of settings

To apply the settings, start the event log trapping service or reload the action definition file for event log trapping by executing the jeveltreload command. For details on the jeveltreload command, see jeveltreload (Windows only) in 15. Commands.

To Page Top

Definition details

An action definition file for event log trapping (ntevent.conf) consists of a destination event server name, retry setting, and one or more filters. Comments are marked with hash marks and disregarded.

server event-server-name

Specify the name of the destination event server for registering JP1 event converted from the event log. Specify a server name that is no more than 255 bytes. Enclose the event server name with double quotation marks. You can only specify an event server that runs on the local host. When no event server is specified, the local host name is assumed.

retry-times retry-count

Specify the number of retries to perform when a connection to the event service fails due to a temporary communication error. Specify a number from 0 to 86400. By default, retry processing is not performed.

retry-interval retry-interval

Specify the retry interval when a connection to the event service fails due to a temporary communication error. This parameter is valid only when you specify a value of 1 or greater in retry-times. The retry interval is the length of time from when the trap fails to connect to the event service until when it next tries to establish connection. This interval does not include the time required for the connection processing. Specify a number from 1 to 600 (seconds). The default is 10.

trap-interval monitoring-interval

Specify the interval over which to monitor the event log. The event log trapping function monitors the event log in real time and also at set intervals. Specify a number from 1 to 180 (seconds). The default is 10.

In Windows Vista and later, this parameter is invalid and will be ignored.

matching-level [0|1]

Specify the comparison level for the event log and definitions when the log entry explanation cannot be read because, for example, you specified a message attribute in the filter condition but the message DLL is not properly configured. When 0 is specified, the next filter condition will be compared skipping the current one. When 1 is specified, the current filter condition is compared. The default is 0.

filter-check-level [0|1]

Specify a checking level when an invalid log type (log type that does not exist in the system) or invalid regular expression is found in a filter condition. Invalidate the filter condition when 0 is specified and the filter condition contains an invalid log type or invalid regular expression. If there are one or more valid filter conditions, the service will start up and the settings will be reloaded successfully. If there are no valid filter conditions, the service will not startup and the settings will not be reloaded. When 1 is specified and one or more of the filter conditions contains an invalid log type or invalid regular expression, the service will not start up and the settings will not be reloaded. The default is 0.

jp1event-send [0|1]

Specify whether to output a message when the event log acquisition fails while monitoring the event log. When 0 is specified, a JP1 event is not output even if the event log acquisition fails. When 1 is specified, a JP1 event (00003A73) is output when the event log acquisition fails. Monitoring might be resumed after the JP1 event indicating failure of event log acquisition. In this case, a JP1 event (00003A74) is output. The default is 0.

Note that a message is output to the integrated trace log regardless of the setting of this parameter. For details on JP1 events, see 17.3 JP1 event details.

In Windows Vista and later, this parameter is invalid and will be ignored.

ext-attr-option extended-attribute-name

Specify this option to create additional extended attributes other than A0 to A6, PLATFORM, and PPNAME. You can specify this parameter in Windows Vista and later only.

You can add multiple extended attributes by separating the attribute names with single-byte spaces. The attributes can be specified in any order.

The following table lists the extended attributes you can specify:

Extended attribute	Meaning
A7	Windows logging level
A8	Windows log keywords
A9	Windows log opcode
OS_VERSION	Windows version number

If you omit this parameter, the event service does not create these extended attributes when it converts JP1 events.

The following shows an example in which all four extended attributes are created:

ext-attr-option A7 A8 A9 OS_VERSION

unicode-trap [0 | 1]

Specify the matching method for event log trapping.

Although the Windows event log is output in the Unicode format, JP1/Base itself does not support Unicode. Therefore, if the event log contains Unicode-specific environment-dependent characters, mismatches in event log trap regular expressions might occur or garbled event log data might be registered in the JP1 event. This parameter causes the event log trapping function to use a Unicode search-based matching method, to prevent mismatches in regular expressions and garbled event log data.

If you specify 0, the matching method for event log trapping is based on the Windows system locale. Because the event log is converted to a character code supported by JP1/Base before matching, mismatches in event log trap regular expressions might occur or garbled event log data might be registered in the JP1 event. Also, the default code set is used when JP1 events are registered.

If you specify 1, the matching method for event log trapping is based on a Unicode search. Because the event log data are matched in its original Unicode-format characters, the event log data can be registered in a JP1 event without garbling. Also, UTF-8 is used as the code set when JP1 events are registered. Extended regular expressions are applied as the regular expressions used for condition statements in event filters.

You can specify this parameter in Windows Vista or later only.

If you omit this parameter, the value 0 is used.

Note that the value set for this parameter cannot be changed by reloading (jeveltreload command). If you change the value set for this parameter, restart the event log trapping service.

To Page Top

Filter syntax

A filter is a set of condition statements for converting event log data into JP1 events. The condition statements within a filter are AND conditions, and those between filters are OR conditions. If you specify multiple filters, conversion is performed when any one of the filters is satisfied. You must specify at least one filter condition. The following figure shows the syntax conventions of a filter.

Figure 16‒7: Filter syntax conventions (action definition file for event log trapping)

To Page Top

Notes

• You can specify a combination of values for the retry count and retry interval that causes the system to continue retrying for more than 24 hours. When retry processing exceeds 24 hours, however, the system aborts retrying and stops the event log trapping service.

• The retry functionality can be used to prevent the Windows media sense functionality from stopping the service.

• When the filter-check-level is set to 0 (or is unspecified) and a filter condition is invalidated, the KAVA3025-W or KAVA3026-W message is output to the event log and integrated trace log. (For file reloading, the message is output only to the standard error output.) Only 10 or fewer messages are output for invalidated filters.

• When the filter-check-level is set to 0 (or is unspecified) and there are no valid filter conditions, the KAVA3027-E or KAVA3028-E message (reloading) is output to the event log and integrated trace log. (For file reloading, the message is output to the event log, integrated trace log, and standard error output.)

• If you distribute an action definition file containing definitions that are valid only in Windows Vista and later to an environment running Windows XP or Windows Server 2003, the definitions are deemed invalid and distribution fails.

• The file name ntevent2.conf is a reserved name. Do not use this name when you back up definition files.

• If 1 is specified for unicode-trap, and a JP1 event for which UTF-8 is used as the code set is registered, upgrade the JP1/Base on the host to which the JP1 event is to be forwarded to version 8 or later.

To Page Top

Supplied action definition file for event log trapping

According to the setting in the supplied action definition file for event log trapping (ntevent.conf), if a connection to the event service fails, the event log trap will retry three times, once per 10-second interval. As conditions for conversion to JP1 events, the defaults also specify that Warning and Error entries output to the System log or Application log are to be converted into JP1 events. The following table shows the settings of the provided file:

retry-times 3
retry-interval 10
 
filter "System"
    type Warning Error
end-filter 
 
filter "Application"
    type Warning Error
end-filter 

If you use the action definition file for event log trapping (ntevent.conf) and forwarding settings file (forward) in their default state, the message KAJP1037-E is output to the event log and converted to a JP1 event when an attempt to forward a JP1 event fails. The converted JP1 event is then resent, and another transfer error will occur.

To prevent the event transfer from looping, change the setting in the action definition file, so that the message KAJP1037-E will not be trapped. A setting example is shown below:

retry-times 3
retry-interval 10
 
filter "System"
type Warning Error
end-filter 
 
# Trap event log entries with severity level Error or Warning
# that were not output by the JP1/Base Event service.
filter "Application"
    type Warning Error
    source !'JP1/Base Event'
end-filter 
 
# Trap event log entries with severity level Error or Warning
# from the JP1/Base Event service, except entries with ID 1037.
filter "Application"
    type Warning Error
    source 'JP1/Base Event'
    id !'1037'
end-filter 

To Page Top

Examples of defining a filter

Definition examples1: Using OR and AND conditions

Definition example using an OR condition

Select data entries of the System log type containing any one of the strings TEXT, MSG, or -W in the explanatory information.

filter "System"
    message 'TEXT' 'MSG' '-W' 
end-filter

Specify an OR condition by separating conditions using spaces and tag characters.

Definition example using an AND condition

Select data entries of the System log type containing all of the strings TEXT, MSG, and -W in the explanatory information.

filter "System"
    message 'TEXT'
    message 'MSG'
    message '-W'
end-filter

Specify an AND condition by separating conditions using a linefeed character. After inserting a linefeed character, write the condition starting from the attribute names.

Definition example 2: Using multiple filters

Trap event log entries that have the Application log type and that satisfy the following conditions.

Filter 1:

• Type: Application log:

• Type: Error

• Explanation: Contains -E and JP1/Base.

Filter 2:

• Type: Application log:

• Type: Warning

• Explanation: Contains -W or warning.

# Filter 1
filter "Application"
    type Error
    message '-E'
    message 'JP1/Base'
end-filter 
# Filter 2
filter "Application"
    type Warning
    message '-W' 'warning'
end-filter 

Definition example 3: Using regular expressions

Trap event log entries that satisfy the following conditions.

• Type: Application log

• Type: Error

• Event ID: 111

• Explanation: Contains -E or MSG, and does not contain TEXT.

filter "Application"
    type Error
    id '^111$'
    message '-E' 'MSG'
    message !'TEXT'
end-filter 

To specify the event ID 111 condition using a regular expression, specify id '^111$'. If you specify id '111', the event ID must contain 111, so event IDs 1112 and 0111 will also satisfy the condition. Writing an exclamation mark in front of the value enclosed with quotation marks selects data that does not match the regular expression. For details on regular expressions, see F. Syntax of Regular Expressions.

Definition example 4: Excluding specific event log entries

Trap event log entries that have System log type and a Warning severity level, but exclude entries that satisfy the following conditions.

• Source: AAA

• Event ID: 111

• Explanation: Contains TEXT.

# Do not trap event log entries from source AAA.
filter "System"
    type Warning
    source !'AAA'
end-filter 
# Trap all event log entries from source AAA, 
# except those with an event ID of 111.
filter "System"
    type Warning
    source 'AAA'
    id !'^111$'
end-filter 
# From source AAA, trap all event log entries 
# whose event ID is 111 and do not contain TEXT 
# in the explanatory information.
filter "System"
    type Warning
    source 'AAA'
    id '^111$'
    message !'TEXT'
end-filter 

To Page Top