Hitachi

JP1 Version 12 JP1/Base User's Guide


Log-file trap startup definition file

Organization of this page

Format

START_OPT=[{SKIP}][<startup-LANG>]monitored-name:jevlogstart-command-options

:

START_OPT_CLS=[(cluster-ID)][<startup-LANG>]monitored-name:jevlogstart-command-options

:

Parameters by type

Required parameters:

None

Custom parameters:
  • START_OPT

  • START_OPT_CLS

File name

jevlog_start.conf

Storage destination directory

In Windows:

installation-folder\conf\event\

In UNIX:

/etc/opt/jp1base/conf/event/

Description

A log-file trap startup definition file specifies the log file traps to be started and stopped when any of the following occurs:

A log-file trap startup definition file is not provided by default. Users can create the file, or the file can be created as definition information (part of a profile) by IM configuration management.

For each log file trap started by a log-file trap startup definition file, a startup record (KAVA3661-I) and startup results (KAVA3662-I) are output to the log-file trap startup execution results log. If a parameter is specified incorrectly, a warning message is output and the parameter is ignored.

Log-file trap startup execution results log data is output to the following files:

In Windows:

installation-folder\log\jevlog_start\jevlog_start{1|2|3}.log

In UNIX:

/var/opt/jp1base/log/jevlog_start/jevlog_start{1|2|3}.log

When a log file trap is started up by using a startup definition file for log file traps, a timeout period applies.

If the jevlogstart command does not finish within the timeout period, processing of the executed jevlogstart command continues, but the startup of the log file trap is interrupted. The command will move on to start up the next definition.

Reference note

Establishing a connection to an event service upon startup of a log file trap might take time, causing a timeout.

If a timeout occurs, the result of executing the jevlogstart command is not output to the "Message" value of the KAVA3662-I message, so you cannot tell whether the startup ended in failure. The log file trap might have stopped due to a failure in connecting to the event service.

It is recommended that you configure the settings to retry connections to event services.

Application of settings

The settings are applied when:

When you execute the jevlogstop command (cluster environment only), the log-file trap management service (daemon) stops the log file traps based on the contents of the log-file trap startup definition file read by the jevlogstart command.

Definition details

The following conventions apply to the contents of the log-file trap startup definition file (jevlog_start.conf):

START_OPT=[{SKIP}][<startup-LANG>]monitored-name:jevlogstart-command-options

Specify the log file traps to start when the log-file trap management service (daemon) starts, and the startup options for each log file trap.

You can specify this parameter and the START_OPT_CLS parameter a maximum of 200 times in total. If the START_OPT parameter and the START_OPT_CLS parameter appear more than 200 times in total, the 201st and subsequent occurrences are ignored.

The log-file trap management service (daemon) does not start or stop the log file traps specified in this parameter when you execute the jevlogstart (cluster environment only) or jevlogstop (cluster environment only) command.

{SKIP}

Specify this option to skip startup of log file trapping by the START_OPT parameter. Specify SKIP in uppercase letters enclosed in curly brackets ({}).

If {SKIP} is omitted, log file trapping is started.

This option is updated when log file trapping is started or stopped by the profile management function of IM Configuration Management. This option is deleted from the definition when log file trapping is started and is added to the definition when log file trapping is stopped.

START_OPT_CLS=[(cluster-ID)][<startup-LANG>]monitored-name:jevlogstart-command-options

Specify the log file traps to start when you execute the jevlogstart (cluster system only) command and stop when you execute the jevlogstop (cluster system only) command, and the startup options for the log file traps. Specify this parameter to collectively start and stop log file traps in the event of a failover in an environment where log files are being monitored on a shared disk in a cluster environment.

You can specify this parameter and the START_OPT parameter a maximum of 200 times in total. If the START_OPT parameter and the START_OPT_CLS parameter appear more than 200 times in total, the 201st and subsequent occurrences are ignored.

The log-file trap management service (daemon) does not start or stop the log file traps specified in this parameter when it starts.

<startup-LANG>

In UNIX, specify the value of the LANG environment variable at execution of the jevlogstart command in angle brackets (< >). For details about the values that can be specified, see Table 3-5 Encodings and LANG environment variable values in UNIX in 3.4.2 Setting the language (for UNIX).

If you omit the value of the LANG environment variable, the assumed encoding is the encoding set in the LANG environment variable used when the log file trap management daemon is started.

In Windows, this parameter has no effect and will be ignored.

monitored-name

Specify the monitor name of the log file trap. The character string preceding the colon (:) is interpreted as the monitor name.

Monitor names must be no more than 30 bytes long, and can contain alphanumeric characters, hyphens, and underscores. Names must start with an alphanumeric character, and are case sensitive.

Do not specify the same monitor name more than once.

jevlogstart-command-options

Specify the options of the jevlogstart command. When the log-file management service (daemon) starts or the jevlogstart (cluster system only) command is executed, the system generates a jevlogstart command line from the contents of this parameter and monitored-name, and uses it to start the log file trap.

For details on the options you can specify, see jevlogstart in 15. Commands.

Only for Windows, there is a limit on the maximum number of specified characters. The maximum number is calculated as follows:

16 + Length of monitored-name in bytes + Length of the options of the jevlogstart command in bytes <= 4,096 bytes

When JP1/Base is linked with IM configuration management, make sure that you specify the name of an action definition file for log file trapping (-f option) for jevlogstart-command-options.

The monitoring target of the log file trap will be the value specified in monitored-name of the START_OPT or START_OPT_CLS parameter, so the -a option is ignored if specified.

If you specify the -cluster option for the jevlogstart (cluster system only) command, the parameter is invalid and the definition is ignored.

The current directory of the jevlogstart command is as follows:

In Windows:

installation-folder\COMMAND

In UNIXL

/opt/jp1base/command

If you specify a relative path in a command option, it is interpreted relative to this location.

(cluster-ID)

Specify the cluster ID in parentheses. The cluster ID is an ID number that identifies a cluster system. In an environment with multiple cluster systems, users can assign any number they choose in the range from 0 to 99.

As the cluster ID, specify a decimal value in the range from 0 to 99. The default is 0.

Defining log file trapping in an environment with multiple cluster systems

In environments with multiple cluster systems, choose a cluster ID for each logical host, and define the log file traps to be started and stopped on each of these hosts when a failover occurs.

The following figure shows how to configure log file trapping in an environment with multiple cluster systems.

Figure 16‒7: Configuration example for environment with multiple cluster systems

[Figure]

In this example, the cluster ID of logical host 1 is 0. The log data output by logical host 1 is monitored by log file trap A and log file trap B. Thus, in the log-file trap startup definition file, define startup options for log file traps A and B for cluster ID 0.

Similarly, define startup options for log file traps C and D as the log file traps for cluster ID 1.

In the cluster software, register the following commands for execution at failover of logical host 1 and logical host 2:

At failover of logical host 1:
  • jevlogstart -cluster 0

  • jevlogstop -cluster 0

At failover of logical host 2:
  • jevlogstart -cluster 1

  • jevlogstop -cluster 1

With these commands registered in this manner, the system starts and ends log file traps A and B when logical host 1 fails over, and starts and ends log file traps C and D when logical host 2 fails over.

For details on the commands you can register with the cluster software, see jevlogstart (cluster environment only) and jevlogstop (cluster environment only) in 15. Commands.

Definition examples

#Specify log file traps to start when service (daemon) starts.
START_OPT=<ja_JP.eucJP>KANSI1:-f /etc/~/jevlog.conf /fil/aaa.log
    :
#Specify log file traps to start when jevlogstart -cluster [cluster-ID] is issued.
#Specify log file traps to stop when jevlogstop -cluster [cluster-ID] is issued.
START_OPT_CLS=(1)<ja_JP.eucJP>KANSI2:-f /etc/~/jevlog.conf /share/aaa.log
    :