factory.ini File

The factory.ini file contains settings for the Job Scheduler. Note that empty settings are ignored and do not overwrite those made in the .

The location of this file can be specified when starting the Job Scheduler using the -ini option. For example,

If the path to the factory.ini file is not specified when calling the Job Scheduler, then the Job Scheduler will attempt to find it by looking according to the following criteria:

This file should be saved under Windows in the folder in which Windows expects to find the .ini files. Normally this is the c:\windows folder. Otherwise the file will be sought in the users' home directory.

Settings

Section: java

Section: job

Specifies whether a task history should be written. If yes, then the Job Scheduler makes an entry in the database for every task carried out. Should the Job Scheduler not be using a database, this entry will be made in a file specified in the -log-dir directory.

This option causes the Job Scheduler to archive and compress the history files with gzip after its last run, should the Job Scheduler be writing the history in files and not in a database.

A task using the Task.set_history_field() method can include fields specified here in the database table.

history_file = file_name Name of the History File (for operation without a database)

Should the parameter history=yes be set and the scheduler not be using a database, then it will make an entry in this file for each task carried out. If no name is specified for the history file, then the Job Scheduler will use its own name. Note that the Job Scheduler will only write a history when the -log-dir option is set.

history_on_process = yes|no|number Write history entry after calling spooler_process()?

Should the Job Job Scheduler be writing a history, then the settings history_on_process=no or =0 cause it to make an entry in the history at the start of a task.

When history_on_process=yes or =1 then Job Scheduler only makes an entry after spooler_process() has been called. This means that no entry will be made, when spooler_open() returns false.

When a number is specified in this setting (history_on_process=x), the Job Scheduler makes an entry in the history only after the x-th execution of spooler_process() .

[ERROR]

SCHEDULER-335

Only "yes", "no" and a number are allowed with ="": error

The Job Scheduler can add the task protocol to the history when the history is being recorded in a database, if required, compressed with gzip. (Here it is important to check that the protocol - which is saved in binary format - can be decompressed after being saved in the database.)

Defines the level with which protocol entries should be written. Every protocol entry is given one of the following categories: error, warn, info, debug1 to debug9 (debug1 is the same as debug).

This setting reduces the numerous e-mails after a job is restarted by way of <delay_after_error> .

This setting only works when <delay_after_error> has been set for as job. Should this not be the case, then mail_on_delay_after_error=all applies.

mail_on_process = yes|no|number Send an e-mail at the start of a task using spooler_process()

Causes the task log to be sent when a task has completed at least the specified number of steps - i.e. calls of spooler_process() . Because non-API tasks do not have steps, the Job Scheduler counts each task as a single step.

mail_queue_dir = directory Directory for e-mails which temporarily cannot be sent

E-mails which cannot be sent (because, for example, the SMTP server cannot be contacted) are stored in this directory.

In order to send these e-mails later it is necessary to write a job which calls up the Mail.dequeue() method.

Section: smtp

mail.smtp.password = password Password for Authentification on an SMTP Server

The password is sent together with the mail.smtp.user entry when registering on an SMTP server in order to be able to send e-mails.

Section: spooler

A simple file name ending in .mdb (e.g. scheduler.mdb) can also be specified here when the Job Scheduler is running on Windows. The Job Scheduler then uses a Microsoft MS Access database of this name, which is located in the protocol directory (see the option -log-dir ). Should such a database not exist, then the Job Scheduler will create this database.

Example:

;    SQL Server 2000 via msbase.jar, msutil.jar, mssqlserver.jar
db = jdbc -class=com.microsoft.jdbc.sqlserver.SQLServerDriver jdbc:microsoft:sqlserver://localhost:1433;selectMethod=Cursor;databaseName=scheduler -user=scheduler -password=secret
;    SQL Server 2000, 2005 via sqljdbc.jar
db = jdbc -class=com.microsoft.sqlserver.jdbc.SQLServerDriver jdbc:sqlserver://localhost:1433;sendStringParametersAsUnicode=false;selectMethod=cursor;databaseName=scheduler -user=scheduler -password=secret

;    MySQL 4.1.7, 5.x
db = jdbc -class=com.mysql.jdbc.Driver jdbc:mysql://localhost/scheduler:3306 -user=scheduler -password=secret

;    Oracle 8.1.7, 9i, 10g
db = jdbc -class=oracle.jdbc.driver.OracleDriver jdbc:oracle:thin:@localhost:1521:orcl -user=scheduler -password=secret

;    PostgreSQL 8.x
db   = jdbc -class=org.postgresql.Driver jdbc:postgresql://localhost:5432/scheduler -user=scheduler -password=secret

;    IBM DB2 8
db   = jdbc -class=com.ibm.db2.jcc.DB2Driver jdbc:db2://localhost:50000/scheduler:driverType=2;retrieveMessagesFromServerOnGetMessage=true; -user=scheduler -password=secret

;    Firebird 1.5
db   = jdbc -class=org.firebirdsql.jdbc.FBDriver jdbc:firebirdsql://localhost:3050/scheduler -user=scheduler -password=secret

;    ODBC
db = odbc -db=scheduler_datasource -user=scheduler -password=secret

;    MS Access Datenbank
db = scheduler.mdb

db_order_step_history_table = name Name of the database table containing the steps in the order history

db_variables_table = name Name of the Job Scheduler internal variable database table

The Job Scheduler records internal counters, for example, the ID of the next free task, in this database table.

This option causes the Job Scheduler to archive and compress the history files with gzip after its last run, should the Job Scheduler be writing the history in files and not in a database.

A task using the Task.set_history_field() method can include fields specified here in the database table.

history_file = file_name Name of the History File (for operation without a database)

history_on_process = yes|no|number Write history entry after calling spooler_process()?

Should the Job Job Scheduler be writing a history, then the settings history_on_process=no or =0 cause it to make an entry in the history at the start of a task.

When history_on_process=yes or =1 then Job Scheduler only makes an entry after spooler_process() has been called. This means that no entry will be made, when spooler_open() returns false.

When a number is specified in this setting (history_on_process=x), the Job Scheduler makes an entry in the history only after the x-th execution of spooler_process() .

[ERROR]

SCHEDULER-335

Only "yes", "no" and a number are allowed with ="": error

The directory in which the HTML files for the Job Scheduler HTTP server are to be found.

Should no entry be made here, then the Job Scheduler uses the html directory in the configuration file directory.

The Job Scheduler only selects elements in the XML configuration whose spooler_id attributes are either empty or set to the value given here.

When the Job Scheduler ID is not specified here, then the Job Scheduler ignores the spooler_id= XML attribute and selects all the elements in the XML configuration.

This setting causes the Job Scheduler to write a detailed protocol. This protocol is intended for use in problem diagnosis. The file name should be fully specified here (i.e. as a full path).

A plus character (+) written directly before the file name causes an already existing protocol to be continued. Otherwise such a protocol will be overwritten.

Categories can be used to extend or restrict the log file. Category names are added (separated by spaces) before the file name, which is then preceded by a larger than (>) sign.

Example:

log = c:/tmp/scheduler.log
log = scheduler.wait >scheduler.log
log = scheduler.wait com_server.* >scheduler.log

log_dir= *stderr allows the Job Scheduler to write log files to the standard output (stderr, normally the screen) .

This setting reduces the numerous e-mails after a job is restarted by way of <delay_after_error> .

This setting only works when <delay_after_error> has been set for as job. Should this not be the case, then mail_on_delay_after_error=all applies.

mail_on_process = yes|no|number Send an e-mail at the start of a task using spooler_process()

mail_on_warning = yes|no An e-mail is sent after a task has ended, should a warning or an error occur

mail_queue_dir = directory Directory for e-mails which temporarily cannot be sent

E-mails which cannot be sent (because, for example, the SMTP server cannot be contacted) are stored in this directory.

In order to send these e-mails later it is necessary to write a job which calls up the Mail.dequeue() method.

max_db_errors = number Number of database errors before the Job Scheduler shuts itself down

The Job Scheduler accepts this number of database errors. In the event of this number of errors being exceeded, and the option need_db=yes, then the scheduler shuts itself down immediately. Should need_db=no then the Job Scheduler will continue without the database.

In the case of need_db= yes, then the errors occurring when attempting to open a database will not be not counted. This means that the Job Scheduler can wait for a database which has not yet been opened.

The Job Scheduler can record the order log in the database - compressed with gzip if required. (Here it is important to check that the log - which is saved in binary format - can be decompressed after being saved in the database.)

On Unix systems /tmp/ is used, should the TMP environment variable not have been set.

1.4 factory.ini File

Settings

Section: java

Section: job

Section: smtp

Section: spooler

Example:

Example: