Configuring the Agent › Agent Parameters in the agentparm.txt File

Agent Parameters in the agentparm.txt File

The agent parameters below appear in the agentparm.txt file in the order listed. The parameter values are set during the agent installation. You can modify these parameters as required. To configure the agent for additional functions, follow the procedures in this chapter.

log.level

Specifies the type and number of logs the agent generates. This parameter is important for troubleshooting.

• 0, 1, or 2—Creates logs for any errors including the receiver and transmitter logs. Level 2 is adequate for production, unless problems arise requiring more details for troubleshooting.
• 3—Adds queues. If this value is specified, the agent ignores the log.maxsize parameter.
• 4 or 5—Adds debugging information. Use log level 5 for setup and initial testing.
• 6-8—Adds tracing information to diagnose a problem. These levels are not intended for continuous use.

Default: 5

Example: log.level=2

log.archive

Defines the log archiving options:

• 0—Appends current date and time to the log file.
• 1—Renames to logfile.archive and starts a new file.
• 2—Removes current file.
• 3—Appends new log entries to the current logs.

Default: 0

log.maxsize=maximum size[B|K|M|G]

Specifies the maximum log size. When the log file exceeds the specified size, the agent archives it and starts a new log file. If the log.archive parameter is set to three, the agent ignores this parameter. The agent does not create an archive file, but it does append all logs. You can specify the following optional modifiers:

• B—Specifies the size in bytes.
• K—Specifies the size in kilobytes.
• M—Specifies the size in megabytes.
• G—Specifies the size in gigabytes.

Note: The default (no modifier) size is in bytes.

Limits: 2G

Default: 1M

agentname

Defines the agent name. You need the agent name when you configure the scheduling manager to work with the agent.

Default: AGENT

Limits: Up to 16 characters; the first character must be a letter

Notes:

• Agent names must begin with an alphabetic character and can contain any alphanumeric characters and the special characters @, $, and underscore (_). Because the scheduling manager uses agent names as file names, use standard file-naming conventions for your operating system.
• For CA Workload Automation DE, the agent name must be in uppercase.

os400.product.library

Specifies the library in which the native i5/OS objects are installed

Default: CAWAGNT113

communication.inputport

Specifies the main port number the agent uses to listen for incoming messages from the scheduling manager. You need this port when you configure the scheduling manager to work with the agent.

Default: 7520

Limits: 1024-65534

Note: On UNIX, ports 1–1023 are reserved ports and require root access.

communication.receiver.socket.main

Specifies the type of socket the agent uses for its main port. The value of this parameter must be different from the communication.receiver.socket.aux parameter. You can specify the following socket types:

• plain
• dylan

Default: plain

Notes:

• CA Workload Automation DE does not require this parameter.
• i5/OS systems use plain socket types only.

communication.managerid_n

Specifies the name of the scheduling manager instance that the agent works with, where n is an integer that corresponds to the scheduling manager being configured.

Default: CENTRAL_MANAGER

Example: MYSERVER

communication.manageraddress_n=address 1;...;address_m

Specifies the address of the scheduling manager that the agent works with, where n is an integer that corresponds to the scheduling manager being configured. This value corresponds to the IP address in the connection details for the scheduling manager. You can specify a list of addresses for the scheduling manager.

Example: 172.24.36.107 (IPv4) or 0:0:0:0:0:FFFF:192.168.00.00 (IPv6)

Notes:

• You can specify a DNS name instead of the IP address for the scheduling manager. However, your agent computer must be able to resolve the DNS name at all times. If there is a DNS outage and your agent computer cannot resolve DNS names, the agent cannot communicate with the scheduling manager.
• If the scheduling manager address never changes, enter the DNS name for the scheduling manager in the hosts file for your agent computer. This entry helps ensure that the IP address can be resolved after DNS disruptions.

communication.managerport_n

Specifies the port that the scheduling manager listens on for communication from agents, where n is an integer that corresponds to the scheduling manager being configured. This value corresponds to the port number in the connection details for the scheduling manager.

Default: 7507

Limits: 1024-65534

communication.monitorobject_n

Specifies the monitor object for the scheduling manager that is used in agent alive ping.

communication.socket_n

Defines the socket type the agent and scheduling manager use for communication, where n is an integer starting at one that corresponds to the scheduling manager being configured. The following socket types are available:

• plain
• dylan

Default: plain

Note: i5/OS systems use plain socket types only.

security.filename

Specifies the path to the security file that contains the security rules that define local security on the agent.

Default: agentinstallDir/security.txt

security.level

Specifies whether local security on the agent is enabled or disabled. Local security on the agent controls which scheduling manager user IDs can perform certain actions, for example, which user IDs can issue CONTROL messages to the agent. If you enable local security, define security rules in a security.txt file.

security.cryptkey

Specifies the path to the text file that stores the encryption key for the agent.

Default: /CA/WA_Agent_R11_3/cyrptkey.txt

initiators.class_n=jobclass,number of initiators

Describes job classes and the number of initiators that can process jobs that are assigned a particular job class. Use a new line for each initiators.class_n parameter, where n is an integer starting at the value 1. By controlling the type and number of initiators, you can have greater control over the initiation of jobs and manually balance the loads on system resources.

The parameter initiators.afmjobclassmap_n relates to this parameter. However, the value of n does not have to match in both parameters.

For UNIX workload, depending on the number of initiators you assign, you may need to increase the number of threads that can be run per process on your operating system.

Examples:

initiators.class_1=Default,1000

initiators.class_2=POJO,100

core.health.monitor.enable

Specifies whether resource usage information within the Java Virtual Machine (JVM), such as memory usage and threads information, should be logged.

• false—Does not log resource usage information within the JVM.
• true—Logs the resource usage information within the JVM to a file called simple_health_monitor.log in the log folder.

Default: true

Note: The log.level parameter must be set to 5 or greater for this information to be logged.

spooldir

Specifies the path to the spool file directory.

Default: spool subdirectory of the agent installation directory

oscomponent.javapath

Specifies the full path to the directory where Java resides.

oscomponent.jvm

Specifies the Java virtual machine (JVM) to use.

plugins.start_internal_n

Specifies the agent plug-in to start by the core Java agent.

n

Denotes an integer assigned to the agent plug-in, starting at 1. The n suffix must increase sequentially for each agent plug-in.

oscomponent.classpath

Specifies the path to jar files required by the agent.

management.snmp.mibfile

Specifies the path to the MIB file that describes the metrics and SNMP traps for the agent.

Default: agentinstalldir/cybermation.mib

management.snmp.host

Identifies the SNMP Manager IP address or DNS name. Your SNMP administrator can provide the host name.

management.snmp.port

Specifies the SNMP Manager UDP port. Your SNMP administrator can provide this port number.

Default: 162

Limits: 1-65535

management.snmp.community

Specifies the type of network the SNMP traps are sent across for SNMP v1 or v2 only. Your SNMP administrator can provide the type.

• public—Identifies an unsecured network, for example, the Internet.
• private—Identifies a secure network, for example, a local area network.

Default: public

ftp.noserver

Specifies whether the agent FTP server is enabled or disabled. If ftp.noserver is set to false, the FTP server is enabled. If the ftp.noserver is set to true, the FTP server is disabled.

Default: true

ftp.serverport

Specifies the port number for the agent to act as an FTP server.

Default: 21

Limits: 1-65534

ftp.client.ssl

Specifies whether all FTP jobs on the agent computer automatically use SSL communication.

• false—Disables SSL communication
• true—Enables SSL communication

ftp.client.ssl.truststore

Specifies the full path name of the truststore file. The default file name is cacerts. You can use keytool, provided with the JRE, to create your own truststore.

ftp.client.ssl.truststore.password

Specifies the encrypted password for the client truststore file, for example cacerts, that contains some common CA X509 certificates.

Default: changeit (encrypted)

Note: You can use the agent password utility to encrypt your password before using it in the agentparm.txt file.

ftp.server.ssl

Specifies that the FTP server handles both non-SSL and SSL FTP.

ftp.server.ssl.keystore.password

Specifies the encrypted password for the server keystore that contains an X509 certificate. This password is sent to the client during the handshake process.

Default: cyberuser (encrypted)

management.connector_n

Identifies the type of management connector the agent uses to connect to an external application, where n is an integer starting from 1. You can specify the following types of connectors:

• JMX—Specifies a JMX connector, built into the agent, that lets you use a JMX console to monitor and control the agent.
• SNMP—Specifies an SNMP connector, built into the agent, that lets you use an SNMP manager to monitor and control the agent.

management.jmx.port

Specifies the port where the JMX connector listens.

Default: 1099

Limits: 1-65534

oscomponent.loginshell

Indicates how to invoke the Shell program when executing a script.

• false—The agent ignores the shell as a login shell.
• true—The agent invokes the shell as a login shell if you specify true. The shell program looks in the directory specified by the HOME environment variable and tries to execute the user’s login scripts (in addition to the .cshrc script).

Default: false

Note: For most systems, this parameter affects only the C and Korn shells. The Bourne shell ignores the oscomponent.loginshell parameter.

oscomponent.defaultshell

Identifies the shell in which scripts are run on the system.

Default: /bin/sh

oscomponent.validshell

Identifies the full path and name of every shell that is valid for use on the agent. Separate each shell with a comma.

Default: /usr/bin/sh,/bin/csh,/bin/ksh,/bin/sh,/bin/bash

Note: This parameter is checked when the oscomponent.checkvalidshell parameter is set to true (the default). If the shell used in a job definition or script is not specified in this parameter, the job fails.

oscomponent.checkvalidshell

Determines whether the agent checks valid shells.

• false—The agent bypasses the valid shell check.
• true—The agent checks valid shells. All shells that jobs use must be specified in the oscomponent.validshell parameter.

Default: true

oscomponent.lookupcommand

Determines how to specify the script or command name to run in a job definition.

• false—The full path to the script or command name must be specified in the job definition.
• true—The script or command name can be specified without the full path in a job definition. The agent looks up the path to the script or command name for the specified user ID.

Default: true

Note: If set to true, ensure that the agent on UNIX is running under the root account.

oscomponent.joblog

Sets whether the agent creates a job log for each UNIX job that runs on the i5/OS system.

• false—Disables job logs
• true—Enables job logs

Default: false

Note: These job logs are different than the job logs the i5/OS system creates.