Autosys Job Management - Unix Installation Guide
Autosys Job Management - Unix Installation Guide
Installation Guide
4.5
This documentation and related computer software program (hereinafter referred to as the “Documentation”) is for
the end user’s informational purposes only and is subject to change or withdrawal by Computer Associates
International, Inc. (“CA”) at any time.
This documentation may not be copied, transferred, reproduced, disclosed or duplicated, in whole or in part,
without the prior written consent of CA. This documentation is proprietary information of CA and protected by the
copyright laws of the United States and international treaties.
Notwithstanding the foregoing, licensed users may print a reasonable number of copies of this documentation for
their own internal use, provided that all CA copyright notices and legends are affixed to each reproduced copy. Only
authorized employees, consultants, or agents of the user who are bound by the confidentiality provisions of the
license for the software are permitted to have access to such copies.
This right to print copies is limited to the period during which the license for the product remains in full force and
effect. Should the license terminate for any reason, it shall be the user’s responsibility to return to CA the reproduced
copies or to certify to CA that same have been destroyed.
To the extent permitted by applicable law, CA provides this documentation “as is” without warranty of any kind,
including without limitation, any implied warranties of merchantability, fitness for a particular purpose or
noninfringement. In no event will CA be liable to the end user or any third party for any loss or damage, direct or
indirect, from the use of this documentation, including without limitation, lost profits, business interruption,
goodwill, or lost data, even if CA is expressly advised of such loss or damage.
The use of any product referenced in this documentation and this documentation is governed by the end user’s
applicable license agreement.
Provided with “Restricted Rights” as set forth in 48 C.F.R. Section 12.212, 48 C.F.R. Sections 52.227-19(c)(1) and (2) or
DFARS Section 252.227-7013(c)(1)(ii) or applicable successor provisions.
Contents iii
Specifying Cross-Instance Job Dependencies ............................................... 1–21
Event Processors and Cross-Dependencies ................................................. 1–22
Event Servers and Cross-Dependencies .................................................... 1–22
High-Availability Options.................................................................... 1–23
Dual Event Servers ...................................................................... 1–23
Running with Dual-Event Servers ..................................................... 1–25
Shadow Event Processor ................................................................. 1–25
Running with a Shadow Event Processor ............................................... 1–27
Related Publications ......................................................................... 1–28
iv Installation Guide
eTrust Client Information ............................................................. 2–13
Unicenter CCI information ............................................................ 2–13
Remote Command Service Information ................................................. 2–13
auto_install Setup ........................................................................ 2–14
System Files Location ................................................................. 2–14
Monitor Type ........................................................................ 2–14
Owner .............................................................................. 2–14
Contents v
Chapter 3: Preparing the Server for Installation
Configuring the Server.........................................................................3–1
Log On to the Server .......................................................................3–1
Installation Directory ......................................................................3–2
Unbundled Sybase.........................................................................4–6
Remote Agent Information .............................................................4–6
vi Installation Guide
Internet Demon ...................................................................... 4–13
Remote Agent Verification ............................................................ 4–13
Contents vii
Chapter 6: Server Installation for Oracle
Before You Begin ..............................................................................6–1
Server Installation .............................................................................6–2
Using the Setup Wizard to Extract Unicenter AutoSys JM from the Media .......................6–4
Oracle Setup Confirmation Utility (recon) .......................................................6–5
Syntax ....................................................................................6–6
Example ..................................................................................6–6
Running the Install Script ......................................................................6–8
Sample Server Installation Script ............................................................6–9
Remote Agent Information .............................................................6–9
Contents ix
Stop the Event Processor ............................................................. 8–19
Run the autobcp Script ............................................................... 8–19
Start the Event Processor ............................................................. 8–21
Installing a Shadow Event Processor .......................................................... 8–23
Notes on Installing a Shadow Event Processor .............................................. 8–23
Installation Steps ........................................................................ 8–24
Restoring the Primary Event Processor .................................................... 8–25
Installing Multiple Instances.................................................................. 8–26
Configuring Multiple Instances Across Platforms ........................................... 8–26
Determining Instance Configuration Settings ........................................... 8–28
Configuring a UNIX Remote Agent for Multiple Instances ............................... 8–29
Configuring Cross-Instance Job Dependencies .............................................. 8–30
Tips on Creating the config.EXTERNAL File............................................ 8–31
x Installation Guide
Appendix B: Platform Notes
DG/UX: TZ Variable Setting ................................................................... B–1
IBM AIX: GUI Display ........................................................................ B–1
Contents xi
Introduction to Unicenter AutoSys
Chapter
1 Job Management
Assumptions
This guide is for system administrators who are responsible for upgrading and
installing Unicenter AutoSys JM as well as configuring the servers and clients
and entering license keys. It assumes familiarity with the UNIX operating system
and with the database server you will be using.
Note: The Unicenter AutoSys JM installation will not let you continue unless
you are running as root on the machine in which you are trying to install
Basics
Unicenter AutoSys JM is an automated job control system for scheduling,
monitoring, and reporting. These jobs can reside on any configured machine that
is attached to a network.
A job is any single command, executable, script, or Windows NT batch file. Each
job definition contains a variety of qualifying attributes, including the conditions
specifying when and where a job should be run.
As with most control systems, many ways exist to correctly define and
implement jobs. It is likely the way you utilize Unicenter AutoSys JM to address
your distributed computing needs will evolve over time. As you become more
familiar with both the features of Unicenter AutoSys JM and the characteristics
of your own jobs, you will also refine your use of Unicenter AutoSys JM.
Before you install and use Unicenter AutoSys JM, however, it is important to
understand the basic system, its components, and how these components work
together.
System Components
The main Unicenter AutoSys JM system components are as follows:
■ Event Processor
■ Remote Agent
In addition, Unicenter AutoSys JM provides utilities to help you define, run, and
maintain instances and jobs. The included utilities are platform-specific;
however, all platforms include the graphical user interface (GUI) and Job
Information Language (JIL). Both the GUI and JIL enable you to define, manage,
monitor, and report on jobs.
NT Client
Remote
Agent
Event Server
(database)
NT Job
Event Server
The event server, or database, is the data repository for all system information
and events as well as all job, monitor, and report definitions. Event server refers
to the database where all the information, events, and job definitions are stored.
Note: The database refers to the specific server instance and the “autosys”
database for that instance. Some utilities, such as isql (Sybase), lets you specify a
particular server and database, and this guide will use the more precise terms of
data server and database in those cases.
You can configure Unicenter AutoSys JM to run using two databases, or dual-
event servers. This feature provides complete redundancy. Therefore, if you lose
one event server due to hardware, software, or network problems, operations
can continue on the second event server without loss of information or
functionality. For more information, see the topic Dual-Event Servers in this
chapter.
Event Processor
The event processor is the heart of Unicenter AutoSys JM; it interprets and
processes all the events it reads from the database. Sometimes called the
event_demon, the event processor is the program, running either as a UNIX
process or as a Windows NT service, which actually runs Unicenter AutoSys JM.
It schedules and starts jobs.
After you start the event processor, the event processor continually scans the
database for events to be processed. When the event processor finds one, it
checks whether the event satisfies the starting conditions for any job in the
database. Based on this information, the event processor first determines what
actions to take, then instructs the appropriate remote agent process to perform
the actions. These actions might be the starting or stopping of jobs, checking for
resources, monitoring existing jobs, or initiating corrective procedures.
You can set up a second event processor, called the shadow event processor. If
the primary event processor fails for some reason, the shadow event processor
will take over the responsibility of interpreting and processing events. For more
information, see the topic Shadow Event Processor in this chapter.
Remote Agent
The remote agent starts the command specified for a given job, sends running
and completion information about a task to the event server, and then exits. If
the remote agent is unable to transfer the information, it waits and tries again
until it can successfully communicate with the database.
The example scenario in the following figure and the numbered explanations
that follow it illustrate the interactions between the event server, the event
processor, and remote agents.
Understanding this example will help you answer many questions that may
arise during your experiences with Unicenter AutoSys JM.
PROCESS PROCESS
PROCESS
PROCESS
5 WorkStation_2
PROCESS Command
Note: In this example, the three primary components are shown running on
different machines. For UNIX, the event processor and the event server typically
run on the same machine.
Explanation
1. From the event server, the event processor reads a new event, which is a
STARTJOB event with a start time condition that has been met. Then the
event processor reads the appropriate job definition from the database and,
based on that definition, determines what action to take. In the example, it
runs the rm /tmp/mystuff/* command on WorkStation_2.
3. The remote agent performs resource checks, such as ensuring that the
minimum specified number of processes are available, then “forks” a child
process that will actually run the specified command.
4. The command completes and exits, and the remote agent captures the
command’s exit code.
5. The remote agent communicates the event (exit code, status, and so forth)
directly to the event server. If the database is unavailable for any reason, the
remote agent will go into a wait and resend cycle until it can deliver the
message.
Only two processes need to be running—the event processor and the event
server. When these two components are running, Unicenter AutoSys JM is fully
operational. The remote agent is started on a client machine once per job. As
soon as the job ends and the remote agent sends a completion event to the
database, the remote agent exits.
Note: The remote agent is started on the client machine by the event processor
talking to the internet demon, inetd, on the client machine. For this to happen,
inetd must also be running. However, because UNIX is responsible for starting
this demon, it is not considered a process.
Interface Components
To define, monitor, and report on jobs, you can use either the GUI or JIL. In
addition, the Operator Console and its dialogs provide a sophisticated method of
monitoring jobs in real time. This feature lets you view all jobs that are defined,
whether or not they are currently active.
■ Multiple Instances
EPLOG
The Unicenter AutoSys JM RCS will send a portion of a Unicenter AutoSys
eplog for a specified instance to the Unicenter AutoSys Web Interface. This
request is processed when a user selects the AutoSys Log option from
Reports Tab of the Unicenter AutoSys Web Interface. The RCS retrieves a
predefined number of lines from the end of the Unicenter AutoSys eplog.
JOBLOG
The JOBLOG request sends the standard output file or standard error file for
a given job to the Unicenter AutoSys JM Web Interface.
CHECKUSER
The CHECKUSER request is used during the Unicenter AutoSys JM Web
Interface Administration AutoSys EDIT Super Sign-on call. The Unicenter
AutoSys JM Web Interface verifies the user credentials by issuing a request
to the RCS to check whether user has the Super Edit privilege.
JIL
The JIL request allows a user to add, remove, or edit a Unicenter AutoSys JM
job by using the Unicenter AutoSys JM Web Interface. When a user edits a
job from the Unicenter AutoSys JM Web Interface Job Management Tab, the
RCS receives the updated job information and uses the JIL program to
update the job in the Event Server.
Machines
From a hardware perspective, the Unicenter AutoSys JM architecture is
composed of the following two types of machines attached to a network:
■ Server machine
The server is the machine on which the event processor or the event server
(database) reside. In a basic UNIX configuration, both the event processor
and the event server reside on the same machine. In a basic Windows NT
configuration, we recommend that the event processor and the event server
reside on separate machines.
■ Client machine
The client is the machine on which the remote agent software resides, and
where jobs run. A remote agent must be installed on the machine with the
event processor, and it can also be installed on separate physical client
machines.
Instance
An instance is one licensed version of Unicenter AutoSys JM software running as
a server with one or more clients, on a single machine or on multiple machines.
You may want to install multiple instances, for example, one instance for
production and another for development. Each instance operates independently
of other AutoSys instances. For information about using multiple instances, see
Multiple Instances in this chapter.
Environment
Access to Unicenter AutoSys JM software and the event server is controlled by
environment variables and the configuration parameters, which must be set for
the event processor and the remote agents to communicate and the commands to
execute. The installation process creates files that are sourced when the user logs
on.
Environment Variables
AUTOSYS
Specifies the full path to the Unicenter AutoSys JM software directory.
AUTOUSER
Specifies the directory containing user configuration files, event processor
output files, archive output files generated during database maintenance,
and sound files (for platforms supporting audio functionality).
AUTOSERV
Specifies the unique, capitalized three-letter name of an instance.
DISPLAY
Referenced to run the AutoSys GUIs.
If only one instance is running, the above variables can be set when a user logs
on by including their definitions in either the .profile or .cshrc file for each user
that accesses Unicenter AutoSys JM.
The installation script generates files that are designed to be sourced by a user
wanting to access Unicenter AutoSys JM. These files can be sourced from a
.profile or .cshrc file. They are found in the $AUTOUSER directory and are listed
following:
These files have the host name appended to the end of their file names so that
client and server environment files cannot be easily confused.
Optionally, frequently used variations for the sendevent, autorep, and eventor
commands can be aliased in the files to be sourced.
Configuration Parameters
For information on the configuration file, see the chapter “Configuring” in the
Unicenter Job Management AutoSys for UNIX User Guide. For information on
the Unicenter AutoSys Administrator, see the chapter “Unicenter AutoSys
Administrator” in the Unicenter AutoSys Job Management for Windows User
Guide.
/etc/auto.profile File
Database Information
Sybase
If you are using a Sybase data server, whether bundled with Unicenter AutoSys
JM or not, the following environment variables are used:
DSQUERY
Defines the name of the Sybase data server.
SYBASE
Specifies the complete path to the Sybase software directory.
The SYBASE environment variable must be set before you run the installation
script. The Sybase software directory contains the Sybase configuration file,
which on UNIX is the interfaces file and on Windows NT is the SQL.INI file.
Unicenter AutoSys JM uses the Sybase configuration file to look up database
information. It is by which the network is navigated to find the Sybase data
server.
Oracle
Directory Structure
The following figure shows the default directory structure for Unicenter AutoSys
JM . This figure refers to the installation directory as AUTOTREE. When this
default structure is used, all the prompts displayed by the install scripts will
match exactly what you see on the screen. The autouser directory can be placed
elsewhere. The new directories are pointed to by the environment variables
AUTOSYS and AUTOUSER.
/usr/vendor/autotree
(bundled DB only)
$AUTOUSER $AUTOSYS $SYBASE
autouser sadb
interfaces file
Configuration Files:
config.$AUTOSERV
autosys.sh.host
autosys.csh.host ASE-12-0 DATA OCS-12-0
etc.
test
bin code dbobj doc sounds snmp
install
Note: These examples use the default values for the environment variables, and
when appropriate, they show both the UNIX and Windows environments.
AutoSys Server
autosys_server
RDBMS
The following figure shows the scenario for connecting to an Oracle database.
AutoSys Server
MYORACLEDB
RDBMS
■ The files named Autosc, Autocons, and Autocal are added to the local app-
defaults directory. This directory is usually one of the following:
/usr/lib/X11/app-defaults
/usr/openwin/lib/app-defaults
Multiple Instances
Multiple instances of Unicenter AutoSys JM can run on the same machine and
can schedule jobs on the same machines without interfering or affecting the other
instances. Each instance uses its own event processor and event server and
operates independently of other instances; however, you can configure instances
to run with cross-instance job dependencies.
You can run multiple instances of Unicenter AutoSys JM on the same network at
the same time. Some of the reasons you may want to run multiple instances are
listed following:
■ Your processing volume is large, and you want to distribute the load down
to the departmental level.
Each instance must have its own event processor specified in the configuration
file and it must have its own instance-specific event servers installed.
Event processors from multiple instances can access the same client machines to
start jobs. To enable this, you must install a remote agent on the client machine
for each instance that will run jobs on that machine.
Different instances can run from the same executables and can have the same
values for the AUTOSYS and AUTOUSER variables, both on the event processor
machine and on client machines. However, each instance must have a unique
value for the AUTOSERV variable and each instance must have its own unique
remote agent port number. You can have unique event server port numbers for
each instance, but this is not required.
The following figure shows two instances of Unicenter AutoSys JM, each with a
single event server. Both instances can send jobs to the same client machine as
long as both instances have a remote agent installed or configured on that client
machine.
OS File OS File
config.ACE File config.PRD File
or NT Registry Settings or NT Registry Settings
Event Event
Server Server
Event Event
Processor Processor
Client
Machine
Note: If you want to install Unicenter AutoSys JM on both UNIX and Windows,
and you want to enable cross-platform job dependencies, then you must supply
different remote agent port numbers for each instance that you install. On
Windows, the remote agent port number must be unique, so if you are running
across UNIX and Windows platforms, you must configure your UNIX
components with the same remote agent port numbers that you are using on the
Windows installations.
For information about installing multiple instances, see the chapter “Advanced
Configurations,” in this guide.
A job defined to run on one instance could have as a starting condition the
successful completion of a job running on a different instance. The specification
for such a job dependency may appear as the following:
condition: success(jobA) AND success(jobB^PRD)
If the dependency specification does not include a caret (^) and a different
instance ID, the current instance will be used by default.
The following figure shows two instances, each with a single event server,
sharing cross-instance dependencies.
Event Event
Server Server
OS File OS File
config.ACE config.PRD File
config.EXTERNAL or NT Registry Settings
Event Event
Processor Processor
Note: If the event server of a target instance is down, the event processor will try
to resend events every five minutes until it can reach the other instance’s event
server.
■ To the ext_job table of the source instance. The entries in this table specify
the status of jobs belonging to other instances in which this instance has
defined job dependencies.
■ To the req_job table of the target instance. The entries in this table specify the
jobs that have a job dependency in a job definition in the source instance.
In both the source instance and target instance tables, jobs are entered using the
job name, a caret symbol (^), and the instance name, like the following:
jobB^PRD
Note: When communicating with event servers, event processors can only
connect to those instances with “like” event servers. That is, instances with
Sybase data servers can only connect with other instances having Sybase data
servers. The same holds true for instances with Oracle and Microsoft SQL Server
databases.
High-Availability Options
Unicenter AutoSys JM provides two high-availability options that lets Unicenter
AutoSys JM keep processing even if an event server or event processor fails due
to hardware or connection problems. These high-availability options are dual-
event servers and a shadow event processor.
You can install and configure the high-availability options at the same time you
install , or you can modify an existing installation to add the high availability
options.
When processing events, the event processor reads from both event servers. If it
detects an event on one server and not the other, it will copy the missing event to
the other server. In this way, a temporary problem in getting events to one of the
servers will not interrupt processing.
In addition, the remote agent sends events and writes to both event servers.
The following figure illustrates a typical configuration running with dual event
servers.
Important! We recommend that the two event servers reside on two different
machines, to prevent a single point of failure.
reads
&
writes
Event Event
Server Server
#1 #2
reads
& writes
writes
Event
Processor
MACHINE A MACHINE B
Start
Job
Remote
Agent
Run Job
User
Command
Remote
MACHINE
When running within dual-event server mode, and the event processor detects
an unrecoverable condition on one of the event servers, it automatically rolls
over to single server mode. A rollover results from one of the following
conditions:
■ The connection to the database is lost, and after the configured number of
reconnect attempts, the database remains unconnected.
If there was an event server rollover on servers running on UNIX, the event
processor edits the $AUTOUSER/config.$AUTOSERV configuration file, on the
server machine only, to comment out the database that has been taken offline.
The event processor makes this change so that utilities attempting to access the
database will know that Unicenter AutoSys JM is now running in single server
mode.
Do not restart the down event server and run in dual-event server mode. Before
starting the event server, you must make sure that the two event servers are
synchronized.
For information on event server recovery, and how to synchronize event servers,
see the chapter “Maintaining ” in the Unicenter AutoSys Job Management for
UNIX User Guide.
Remote
Agent
.dibs file
Third
MACHINE
reads
&
writes
Event Event
Server Server
#1 #2
reads
& writes
writes
Primary Shadow
Processor Processor
PING
MACHINE A MACHINE B
Start
Job
Remote
Agent
Run Job
User
Command
Client
MACHINE
The shadow event processor is normally in idle mode, listening for routine pings
from the primary event processor, which indicate all is well. If the shadow event
processor stops receiving this signal, it assumes the primary event processor has
failed.
If the shadow event processor does not receive the signal, it checks the third
machine (defined in the configuration file) for the .dibs file. If it cannot connect
to the third machine, the shadow event processor shuts down. If it can connect
and cannot locate the .dibs file, the shadow event processor creates the file,
attempts to signal the primary event processor to stop, and takes over processing
the events. If the file already exists, it shuts down.
Similarly, if the primary event processor cannot locate and signal the shadow
event processor, the primary processor checks the third machine for the .dibs file
and follows the same procedure as the shadow event processor (described
previously).
The shadow event processor is designed primarily for the situation where the
machine on which the primary event processor runs goes down, or the network
on which this processor runs goes down. Particular care is given to ensuring that
both event processors never take over at the same time. To achieve this,
Unicenter AutoSys JM uses the third machine and the existence of the .dibs file
to resolve contentions and to eliminate the case where one processor takes over
because its own network is down.
The shadow event processor is not guaranteed to take over in 100% of the cases
where it theoretically could. For example, in the case of network problems,
Unicenter AutoSys JM may not be able to determine which event processor is the
functional one. In this case, both processors will shut down.
For more information about event processor rollover and recovery, see the
chapter “Maintaining” in the Unicenter AutoSys Job Management for UNIX User
Guide.
Related Publications
In addition to this Installation Guide, the following documentation is provided:
■ Unicenter AutoSys Job Management for UNIX User Guide, which describes
how to use Unicenter AutoSys JM to define, run, manage, monitor, and
report on jobs. In addition, it describes how to run and manage the
databases, and it describes security and the configuration file.
Basics
Unicenter AutoSys JM is an automated job control system for scheduling,
monitoring, and reporting. These jobs can reside on any configured machine that
is attached to a network.
A job is any single command, executable, script, or Windows NT batch file. Each
job definition contains a variety of qualifying attributes, including the conditions
specifying when and where a job should be run.
As with most control systems, many ways exist to correctly define and
implement jobs. It is likely the way you utilize Unicenter AutoSys JM to address
your distributed computing needs will evolve over time. As you become more
familiar with both the features of Unicenter AutoSys JM and the characteristics
of your own jobs, you will also refine your use of Unicenter AutoSys JM.
Before you install and use Unicenter AutoSys JM, however, it is important to
understand the basic system, its components, and how these components work
together.
System Components
The main Unicenter AutoSys JM system components are as follows:
■ Event Processor
■ Remote Agent
In addition, Unicenter AutoSys JM provides utilities to help you define, run, and
maintain instances and jobs. The included utilities are platform-specific;
however, all platforms include the graphical user interface (GUI) and Job
Information Language (JIL). Both the GUI and JIL enable you to define, manage,
monitor, and report on jobs.
NT Client
Remote
Agent
Event Server
(database)
NT Job
Event Server
The event server, or database, is the data repository for all system information
and events as well as all job, monitor, and report definitions. Event server refers
to the database where all the information, events, and job definitions are stored.
Note: The database refers to the specific server instance and the “autosys”
database for that instance. Some utilities, such as isql (Sybase), lets you specify a
particular server and database, and this guide will use the more precise terms of
data server and database in those cases.
You can configure Unicenter AutoSys JM to run using two databases, or dual-
event servers. This feature provides complete redundancy. Therefore, if you lose
one event server due to hardware, software, or network problems, operations
can continue on the second event server without loss of information or
functionality. For more information, see the topic Dual-Event Servers in this
chapter.
Event Processor
The event processor is the heart of Unicenter AutoSys JM; it interprets and
processes all the events it reads from the database. Sometimes called the
event_demon, the event processor is the program, running either as a UNIX
process or as a Windows NT service, which actually runs Unicenter AutoSys JM.
It schedules and starts jobs.
After you start the event processor, the event processor continually scans the
database for events to be processed. When the event processor finds one, it
checks whether the event satisfies the starting conditions for any job in the
database. Based on this information, the event processor first determines what
actions to take, then instructs the appropriate remote agent process to perform
the actions. These actions might be the starting or stopping of jobs, checking for
resources, monitoring existing jobs, or initiating corrective procedures.
You can set up a second event processor, called the shadow event processor. If
the primary event processor fails for some reason, the shadow event processor
will take over the responsibility of interpreting and processing events. For more
information, see the topic Shadow Event Processor in this chapter.
Remote Agent
The remote agent starts the command specified for a given job, sends running
and completion information about a task to the event server, and then exits. If
the remote agent is unable to transfer the information, it waits and tries again
until it can successfully communicate with the database.
The example scenario in the following figure and the numbered explanations
that follow it illustrate the interactions between the event server, the event
processor, and remote agents.
Understanding this example will help you answer many questions that may
arise during your experiences with Unicenter AutoSys JM.
PROCESS PROCESS
PROCESS
PROCESS
5 WorkStation_2
PROCESS Command
Note: In this example, the three primary components are shown running on
different machines. For UNIX, the event processor and the event server typically
run on the same machine.
Explanation
1. From the event server, the event processor reads a new event, which is a
STARTJOB event with a start time condition that has been met. Then the
event processor reads the appropriate job definition from the database and,
based on that definition, determines what action to take. In the example, it
runs the rm /tmp/mystuff/* command on WorkStation_2.
3. The remote agent performs resource checks, such as ensuring that the
minimum specified number of processes are available, then “forks” a child
process that will actually run the specified command.
4. The command completes and exits, and the remote agent captures the
command’s exit code.
5. The remote agent communicates the event (exit code, status, and so forth)
directly to the event server. If the database is unavailable for any reason, the
remote agent will go into a wait and resend cycle until it can deliver the
message.
Only two processes need to be running—the event processor and the event
server. When these two components are running, Unicenter AutoSys JM is fully
operational. The remote agent is started on a client machine once per job. As
soon as the job ends and the remote agent sends a completion event to the
database, the remote agent exits.
Note: The remote agent is started on the client machine by the event processor
talking to the internet demon, inetd, on the client machine. For this to happen,
inetd must also be running. However, because UNIX is responsible for starting
this demon, it is not considered a process.
Interface Components
To define, monitor, and report on jobs, you can use either the GUI or JIL. In
addition, the Operator Console and its dialogs provide a sophisticated method of
monitoring jobs in real time. This feature lets you view all jobs that are defined,
whether or not they are currently active.
■ Multiple Instances
EPLOG
The Unicenter AutoSys JM RCS will send a portion of a Unicenter AutoSys
eplog for a specified instance to the Unicenter AutoSys Web Interface. This
request is processed when a user selects the AutoSys Log option from
Reports Tab of the Unicenter AutoSys Web Interface. The RCS retrieves a
predefined number of lines from the end of the Unicenter AutoSys eplog.
JOBLOG
The JOBLOG request sends the standard output file or standard error file for
a given job to the Unicenter AutoSys JM Web Interface.
CHECKUSER
The CHECKUSER request is used during the Unicenter AutoSys JM Web
Interface Administration AutoSys EDIT Super Sign-on call. The Unicenter
AutoSys JM Web Interface verifies the user credentials by issuing a request
to the RCS to check whether user has the Super Edit privilege.
JIL
The JIL request allows a user to add, remove, or edit a Unicenter AutoSys JM
job by using the Unicenter AutoSys JM Web Interface. When a user edits a
job from the Unicenter AutoSys JM Web Interface Job Management Tab, the
RCS receives the updated job information and uses the JIL program to
update the job in the Event Server.
Machines
From a hardware perspective, the Unicenter AutoSys JM architecture is
composed of the following two types of machines attached to a network:
■ Server machine
The server is the machine on which the event processor or the event server
(database) reside. In a basic UNIX configuration, both the event processor
and the event server reside on the same machine. In a basic Windows NT
configuration, we recommend that the event processor and the event server
reside on separate machines.
■ Client machine
The client is the machine on which the remote agent software resides, and
where jobs run. A remote agent must be installed on the machine with the
event processor, and it can also be installed on separate physical client
machines.
Instance
An instance is one licensed version of Unicenter AutoSys JM software running as
a server with one or more clients, on a single machine or on multiple machines.
You may want to install multiple instances, for example, one instance for
production and another for development. Each instance operates independently
of other AutoSys instances. For information about using multiple instances, see
Multiple Instances in this chapter.
Environment
Access to Unicenter AutoSys JM software and the event server is controlled by
environment variables and the configuration parameters, which must be set for
the event processor and the remote agents to communicate and the commands to
execute. The installation process creates files that are sourced when the user logs
on.
Environment Variables
AUTOSYS
Specifies the full path to the Unicenter AutoSys JM software directory.
AUTOUSER
Specifies the directory containing user configuration files, event processor
output files, archive output files generated during database maintenance,
and sound files (for platforms supporting audio functionality).
AUTOSERV
Specifies the unique, capitalized three-letter name of an instance.
DISPLAY
Referenced to run the AutoSys GUIs.
If only one instance is running, the above variables can be set when a user logs
on by including their definitions in either the .profile or .cshrc file for each user
that accesses Unicenter AutoSys JM.
The installation script generates files that are designed to be sourced by a user
wanting to access Unicenter AutoSys JM. These files can be sourced from a
.profile or .cshrc file. They are found in the $AUTOUSER directory and are listed
following:
These files have the host name appended to the end of their file names so that
client and server environment files cannot be easily confused.
Optionally, frequently used variations for the sendevent, autorep, and eventor
commands can be aliased in the files to be sourced.
Configuration Parameters
For information on the configuration file, see the chapter “Configuring” in the
Unicenter Job Management AutoSys for UNIX User Guide. For information on
the Unicenter AutoSys Administrator, see the chapter “Unicenter AutoSys
Administrator” in the Unicenter AutoSys Job Management for Windows User
Guide.
/etc/auto.profile File
Database Information
Sybase
If you are using a Sybase data server, whether bundled with Unicenter AutoSys
JM or not, the following environment variables are used:
DSQUERY
Defines the name of the Sybase data server.
SYBASE
Specifies the complete path to the Sybase software directory.
The SYBASE environment variable must be set before you run the installation
script. The Sybase software directory contains the Sybase configuration file,
which on UNIX is the interfaces file and on Windows NT is the SQL.INI file.
Unicenter AutoSys JM uses the Sybase configuration file to look up database
information. It is by which the network is navigated to find the Sybase data
server.
Oracle
Directory Structure
The following figure shows the default directory structure for Unicenter AutoSys
JM . This figure refers to the installation directory as AUTOTREE. When this
default structure is used, all the prompts displayed by the install scripts will
match exactly what you see on the screen. The autouser directory can be placed
elsewhere. The new directories are pointed to by the environment variables
AUTOSYS and AUTOUSER.
/usr/vendor/autotree
(bundled DB only)
$AUTOUSER $AUTOSYS $SYBASE
autouser sadb
interfaces file
Configuration Files:
config.$AUTOSERV
autosys.sh.host
autosys.csh.host ASE-12-0 DATA OCS-12-0
etc.
test
bin code dbobj doc sounds snmp
install
Note: These examples use the default values for the environment variables, and
when appropriate, they show both the UNIX and Windows environments.
AutoSys Server
autosys_server
RDBMS
The following figure shows the scenario for connecting to an Oracle database.
AutoSys Server
MYORACLEDB
RDBMS
■ The files named Autosc, Autocons, and Autocal are added to the local app-
defaults directory. This directory is usually one of the following:
/usr/lib/X11/app-defaults
/usr/openwin/lib/app-defaults
Multiple Instances
Multiple instances of Unicenter AutoSys JM can run on the same machine and
can schedule jobs on the same machines without interfering or affecting the other
instances. Each instance uses its own event processor and event server and
operates independently of other instances; however, you can configure instances
to run with cross-instance job dependencies.
You can run multiple instances of Unicenter AutoSys JM on the same network at
the same time. Some of the reasons you may want to run multiple instances are
listed following:
■ Your processing volume is large, and you want to distribute the load down
to the departmental level.
Each instance must have its own event processor specified in the configuration
file and it must have its own instance-specific event servers installed.
Event processors from multiple instances can access the same client machines to
start jobs. To enable this, you must install a remote agent on the client machine
for each instance that will run jobs on that machine.
Different instances can run from the same executables and can have the same
values for the AUTOSYS and AUTOUSER variables, both on the event processor
machine and on client machines. However, each instance must have a unique
value for the AUTOSERV variable and each instance must have its own unique
remote agent port number. You can have unique event server port numbers for
each instance, but this is not required.
The following figure shows two instances of Unicenter AutoSys JM, each with a
single event server. Both instances can send jobs to the same client machine as
long as both instances have a remote agent installed or configured on that client
machine.
OS File OS File
config.ACE File config.PRD File
or NT Registry Settings or NT Registry Settings
Event Event
Server Server
Event Event
Processor Processor
Client
Machine
Note: If you want to install Unicenter AutoSys JM on both UNIX and Windows,
and you want to enable cross-platform job dependencies, then you must supply
different remote agent port numbers for each instance that you install. On
Windows, the remote agent port number must be unique, so if you are running
across UNIX and Windows platforms, you must configure your UNIX
components with the same remote agent port numbers that you are using on the
Windows installations.
For information about installing multiple instances, see the chapter “Advanced
Configurations,” in this guide.
A job defined to run on one instance could have as a starting condition the
successful completion of a job running on a different instance. The specification
for such a job dependency may appear as the following:
condition: success(jobA) AND success(jobB^PRD)
If the dependency specification does not include a caret (^) and a different
instance ID, the current instance will be used by default.
The following figure shows two instances, each with a single event server,
sharing cross-instance dependencies.
Event Event
Server Server
OS File OS File
config.ACE config.PRD File
config.EXTERNAL or NT Registry Settings
Event Event
Processor Processor
Note: If the event server of a target instance is down, the event processor will try
to resend events every five minutes until it can reach the other instance’s event
server.
■ To the ext_job table of the source instance. The entries in this table specify
the status of jobs belonging to other instances in which this instance has
defined job dependencies.
■ To the req_job table of the target instance. The entries in this table specify the
jobs that have a job dependency in a job definition in the source instance.
In both the source instance and target instance tables, jobs are entered using the
job name, a caret symbol (^), and the instance name, like the following:
jobB^PRD
Note: When communicating with event servers, event processors can only
connect to those instances with “like” event servers. That is, instances with
Sybase data servers can only connect with other instances having Sybase data
servers. The same holds true for instances with Oracle and Microsoft SQL Server
databases.
High-Availability Options
Unicenter AutoSys JM provides two high-availability options that lets Unicenter
AutoSys JM keep processing even if an event server or event processor fails due
to hardware or connection problems. These high-availability options are dual-
event servers and a shadow event processor.
You can install and configure the high-availability options at the same time you
install , or you can modify an existing installation to add the high availability
options.
When processing events, the event processor reads from both event servers. If it
detects an event on one server and not the other, it will copy the missing event to
the other server. In this way, a temporary problem in getting events to one of the
servers will not interrupt processing.
In addition, the remote agent sends events and writes to both event servers.
The following figure illustrates a typical configuration running with dual event
servers.
Important! We recommend that the two event servers reside on two different
machines, to prevent a single point of failure.
reads
&
writes
Event Event
Server Server
#1 #2
reads
& writes
writes
Event
Processor
MACHINE A MACHINE B
Start
Job
Remote
Agent
Run Job
User
Command
Remote
MACHINE
When running within dual-event server mode, and the event processor detects
an unrecoverable condition on one of the event servers, it automatically rolls
over to single server mode. A rollover results from one of the following
conditions:
■ The connection to the database is lost, and after the configured number of
reconnect attempts, the database remains unconnected.
If there was an event server rollover on servers running on UNIX, the event
processor edits the $AUTOUSER/config.$AUTOSERV configuration file, on the
server machine only, to comment out the database that has been taken offline.
The event processor makes this change so that utilities attempting to access the
database will know that Unicenter AutoSys JM is now running in single server
mode.
Do not restart the down event server and run in dual-event server mode. Before
starting the event server, you must make sure that the two event servers are
synchronized.
For information on event server recovery, and how to synchronize event servers,
see the chapter “Maintaining ” in the Unicenter AutoSys Job Management for
UNIX User Guide.
Remote
Agent
.dibs file
Third
MACHINE
reads
&
writes
Event Event
Server Server
#1 #2
reads
& writes
writes
Primary Shadow
Processor Processor
PING
MACHINE A MACHINE B
Start
Job
Remote
Agent
Run Job
User
Command
Client
MACHINE
The shadow event processor is normally in idle mode, listening for routine pings
from the primary event processor, which indicate all is well. If the shadow event
processor stops receiving this signal, it assumes the primary event processor has
failed.
If the shadow event processor does not receive the signal, it checks the third
machine (defined in the configuration file) for the .dibs file. If it cannot connect
to the third machine, the shadow event processor shuts down. If it can connect
and cannot locate the .dibs file, the shadow event processor creates the file,
attempts to signal the primary event processor to stop, and takes over processing
the events. If the file already exists, it shuts down.
Similarly, if the primary event processor cannot locate and signal the shadow
event processor, the primary processor checks the third machine for the .dibs file
and follows the same procedure as the shadow event processor (described
previously).
The shadow event processor is designed primarily for the situation where the
machine on which the primary event processor runs goes down, or the network
on which this processor runs goes down. Particular care is given to ensuring that
both event processors never take over at the same time. To achieve this,
Unicenter AutoSys JM uses the third machine and the existence of the .dibs file
to resolve contentions and to eliminate the case where one processor takes over
because its own network is down.
The shadow event processor is not guaranteed to take over in 100% of the cases
where it theoretically could. For example, in the case of network problems,
Unicenter AutoSys JM may not be able to determine which event processor is the
functional one. In this case, both processors will shut down.
For more information about event processor rollover and recovery, see the
chapter “Maintaining” in the Unicenter AutoSys Job Management for UNIX User
Guide.
Related Publications
In addition to this Installation Guide, the following documentation is provided:
■ Unicenter AutoSys Job Management for UNIX User Guide, which describes
how to use Unicenter AutoSys JM to define, run, manage, monitor, and
report on jobs. In addition, it describes how to run and manage the
databases, and it describes security and the configuration file.
Memory
Disk Space
The total disk space requirement to install Unicenter AutoSys JM with a bundled
Sybase database on the server is 600 MB. To install only the application (without
bundled Sybase), 150 MB of available disk space is required. To install the Oracle
application binaries, 350 MB of disk space is required.
Oracle Requirements
Sybase Requirements
The hardware and software requirements for the Sybase Adaptive Server
Enterprise Version 12.0 are listed as following.
92 MB (64-bit)
The following operating system patches are required for the Adaptive Server on
HP-UX 11.0 to run Adaptive Server 12.0 components for both 32-bit and 64-bit
versions:
■ PHKL_17091 – PM/VM/UFS/async/scsi/io/DMAPI/p2p_bcopy/MPI
AIX 4.3.2 requires the following operating system patches to run Adaptive
Server 12.0 components:
■ TY01642
■ TY02407
Solaris 2.6 (32-bit) requires the following operating system patches to run
Adaptive Server 12.0 components:
■ 105284-16
■ 105181-06
■ 105210-09
■ 105529-03
■ 105786-05
■ 104668-09 (c 4.2)
Solaris 2.7 (64-bit) requires the following operating system patches to run
Adaptive Server 12.0 components:
■ 106541-03
■ 106327-05
■ 106300-06
The required database sizes are the minimum recommended sizes. If your job
load is large, you should create a larger database. The size requirements for your
database depend on the following criteria:
3. How often the jobs are run. (Every time a job runs, it generates at least
three events and an entry in the job_runs table.)
4. How often the database is cleaned. (The utility called dbmaint handles
this; see the chapter “Maintaining” in the Unicenter AutoSys Job
Management for UNIX User Guide.)
Selecting Components
The installation lets you install the following listed software components:
Server
The server installation sets up the database and various configuration files,
and then configures the server machine to run as a client as well. This
enables you to run jobs on a server machine.
Client
The client installation must be performed on every machine that you will use
to run, monitor, or define jobs.
With the Unicenter AutoSys JM 4.5 base product, the Java Listener which is
currently shipped on the Unicenter AutoSys JM Web Interface CD has been
updated to run as a “C” based application. This new “C” based application is
now known as the Remote Command Service (RCS) and is included on the
Unicenter AutoSys JM 4.5 base product CD. Having written the RCS as a
portable “C” based application it eliminates the need to have to install the Java
Runtime Environment (JRE) on the Unicenter AutoSys JM Event Processor
machine or Remote Agent machines to which the Unicenter AutoSys JM Web
Interface was to access data from. The following sections describe each
component in some addition detail.
RCS component should be selected for installation when installing the Unicenter
AutoSys JM 4.5 base product if you plan to utilize the Unicenter AutoSys JM 4.5
Web Interface. The Unicenter AutoSys JM 4.5 Web Interface requires RCS to be
installed in order to service requests initiated on behalf of the Web Interface
when running against Unicenter AutoSys JM 4.5.
RCS is included as part of the base installation with the following platforms.
The Java Listener component should be selected for installation if installing the
Unicenter AutoSys JM Web Interface and the Web Interface will be initiating
requests against earlier versions of Unicenter AutoSys JM (For example; 4.0).
For more information on requirement and details on installing the Java Listener,
see the Unicenter AutoSys Job Management Web Interface Installation Guide.
Listener Compatibility
The following table illustrates which listener to install when interfacing the
Unicenter AutoSys JM Web Interface with the Unicenter AutoSys JM base
product.
Notes:
■ The 4.0 Java Listener is included on the Unicenter AutoSys JM Web Interface
4.0 CD.
■ The 4.5 Java Listener is included on the Unicenter AutoSys JM Web Interface
4.5 CD.
Identifying Machines
Before you install, identify the machines on which you will install the required
components.
Server Machines
The server machine is a machine on which the database, the event processor, or
both reside.
You must perform the modifications described in the chapter “Preparing the
Server for Installation” in this guide, before you perform the installation.
Identify one reliable machine on which to install the database. To ensure high
availability of the database, you can install dual-event servers; in which case,
you identify two reliable machines on which to install databases. The terms
event server and database are often used interchangeably.
You can also install a shadow event processor to ensure high availability of the
event processor. This requires that you identify two additional machines—a
shadow machine and a third machine. The primary and shadow event processor
machines and the third machine must all be of the same type, either Windows or
UNIX. All three machines must be defined by the same instance.
Client Machines
Identify one or more machines on which to install the remote agent. You can
define a client machine to run jobs only, or to run both jobs and the GUIs for
defining and monitoring jobs.
To avoid intrusion, eTrust AC is particular about the spelling of user names and
host names. This affects Unicenter AutoSys JM users in the following
circumstances:
In the first case, the eTrust server uses the IP address of the Policy Manager
machine to find its host name. It then checks that the user making the request
has administrative authority over the PMDB from that host. In other words, the
Policy Manager user must be defined as an administrator of the PMDB, and
must have write access to the TERMINAL whose name matches the name of the
host where the Policy Manager is running.
In the second case, the subscriber uses a similar procedure to authenticate the
PMD server. It uses the server's IP address to find the server's host name. It
checks that the server's host name matches the subscriber's expected PMD server
name and that the user who changed the policy is an administrator of the
subscriber.
4. Make sure the Show Connection Dialog box is checked under General
Settings.
5. Click OK.
When entering a Windows user name with a domain name into a UNIX
machine, double the backslash, for example WORKGROUP\\Administrator.
UNIX systems are sensitive to upper and lower case letters, even in the names of
Windows users.
Use the PMD server to find the name of a Policy Manager host. Use a subscriber
to find the name of the PMD server. Remember to do this on the machine
receiving the connection, not the machine whose name you want to know how to
spell.
Beware that different subscribers on your network might use different name
services, and therefore might require different spellings of the PMD server name.
Some might require the domain name and others might require you to omit it.
Try it without the domain name, then with it. Substitute the host's IP address for
ipaddr where it appears.
Look for the words files, nis, nisplus, and dns in the output. Use the
commands to the right of the $ signs in steps 2 and 3 in the order the
corresponding name services appear in nsswitch.conf. Stop when you find a
match. If you don't have the nsswitch.conf file, try all of the commands.
Examples:
$ grep '^hosts' /etc/nsswitch.conf
hosts: files nis dns
$ fgrep -i morpheus /etc/hosts
$ ypmatch morpheus hosts
192.168.2.25 morpheus hqsun56
$ fgrep 192.168.2.25 /etc/hosts
192.168.2.25 hqsrv56
In the previous example, the correct spelling is hqsrv56, since that's the first
name encountered for that IP address.
$ grep '^hosts' /etc/nsswitch.conf
hosts: files dns
$ fgrep -i morpheus /etc/hosts
$ nslookup morpheus
Server: hqsrv2.com.biz
Address: 192.168.1.7
Name: hqsrv56.com.biz
Address: 192.168.2.25
Aliases: morpheus.com.biz
$ fgrep 192.168.2.25 /etc/hosts
$ nslookup 192.168.2.25
Server: hqsrv2.com.biz
Address: 192.168.1.7
Name: hqsrv56.com.biz
Address: 192.168.2.25
In the previous examples, this subscriber doesn't use NIS, and doesn't have the
PMD server in its hosts file. The correct spelling in this case is hqsrv56.com.biz.
Wizard Setup
Prior to installation, you will need the following information for the setup
wizard.
This information is for the eTrust server which will contain the pmdb.
Admin. Users
List the OS users who will have permission to administer the eTrust AC on
Admin Hosts
Admin. Hosts
A list of hosts from which Admin Users may administer the pmdb.
Subscribers
The host names of eTrust AC clients which will subscribe to this pmdb.
This information is for the eTrust client which will subscribe to the pmdb.
Admin. Users
A list of users who will have permission to administer the eTrust AC on this
host.
Note: In the eTrust client (seosdb), you must specify an admin user that is
defined in the eTrust server (pmdb). If this is not done then client (seosdb) will
not be synchronized with the server (pmdb).
Server host
Installation directory
Remote Hosts
Installation directory
Port Number
Valid Hosts
The host names of machine which will communicate with the RCS.
Directories
The fully qualified names of directories where the RCS will look for standard
output and standard error files from jobs.
auto_install Setup
After the Setup Wizard, you will need the following information for the
auto_install
Specify the location of your system files, either local, NIS, or NIS+. If your
system uses NIS or NIS+, see NIS or NIS+ Only: Configure Internet Demon
Services in the chapter “Preparing the Server for Installation,” in this guide.
Monitor Type
Owner
A UNIX user will be assigned ownership of the software files. This can be any
user except root. If this user does not exist, create it now. This user’s home
directory can be any directory except one of those found under the standard
installation directory autotree.
The owner has no special permissions regarding jobs unless given Superuser
permissions during the installation process.
Instance Information
The $AUTOUSER directory, which contains configuration files, output files, and
user-defined sound files, can be placed in a different directory than the standard
installation directory (autotree). By convention, the $AUTOUSER directory is
located in autotree/autouser.
The logical name of the database (event server). When performing database
queries, Unicenter AutoSys JM uses this value to locate the event server.
If you are installing with bundled Sybase, the name of the data server is
AUTOSYSDB.
The host name of the machine on which the event server for the instance is
installed and will run.
The name of the database. For bundled Sybase, this name is autosys.
SA User (Sybase)
The user who was granted the System Administrator and System Security
Officer (SSO) roles for the database.
SA Password (Sybase)
The password for the user who was granted the System Administrator (SA) and
System Security Officer (SSO) roles for the database.
Bundled Sybase
Note: Bundled Sybase users cannot change the size of the database shipped with
the software.
Unbundled Sybase
The steps for creating and configuring the Sybase database are detailed in
Configuring an Existing Unbundled Sybase Database in the chapter, “Preparing
the Server for Installation” in this guide.
The Net8/Oracle Net TNS alias for the event server that will contain the
database. AutoSys requires that Net8/Oracle Net be installed on the database
machine. The TNS alias name must be configured in the Oracle tnsnames.ora
configuration file.
The system TNS configuration file is tnsnames.ora. You must specify the path to
the tnsnames.ora file.
The name of the tablespace on event server that will contain the database tables.
The recommended size is 100 MB.
The name of the tablespace on the event server that will contain the database
indexes. The recommended size is 40 MB.
The name of the tablespace on the event server that will contain the temporary
tablespace.
The name of the user who was granted the DBA role for the database.
The password of the user who was granted the DBA role for the database.
The steps for configuring the Oracle database are detailed in Configuring an
Existing Oracle Database in the chapter, “Preparing the Server for Installation”
in this guide.
The recon utility is provided to verify that your Oracle environment is properly
configured. As described in the Oracle Setup Utility in the chapter, “Server
Installation for Oracle” in this guide.
The path to the remote agent executable. The remote agent executable can be
installed in any valid directory. The default for server installations is
$AUTOSYS/bin; for clients, it is /usr/local/bin. Because the actual path to the
executable is written in the inetd configuration file, you should be cautious
about changing it. Therefore, choose a stable directory that is not subject to
removal.
You must specify the path and file name of the configuration file for the internet
services demon (inetd). Typically, this file is:
/etc/inetd.conf;
The default port number for the remote agent is 5280. If this number is already in
use, or if you wish to install multiple versions of the remote agent, you can
change this number.
Admin. Hosts
Subscribers
Server host
Installation Directory
Remote Hosts
Installation Directory
Port Number
Valid Hosts
Directories
Admin. Hosts
Subscribers
Server host
Installation Directory
Port Number
Valid Hosts
Directories
Server Checklist
Server Checklist
Item Description
Platform
Operating system/version
Host name
Host ID
Password of root
Server Checklist
Item Description
Name of data server that will contain
database (DSQUERY)
Value of ORACLE_HOME
Value of TNS_ADMIN
Server Checklist
Item Description
Name of the database user who has
been granted DBA role
Client Checklist
Use the following table to define your client. Entries with two stars (**) must
have the same response as the corresponding entry in the Server checklist.
Client Checklist
Item Description
Platform
Operating system/version
Host name
Host ID
Password of root
Client Checklist
Item Description
Password of database user who has
been granted both SA and SSO roles
Value of ORACLE_HOME
Value of TNS_ADMIN
3 Installation
This chapter describes how to configure your server machine and its operating
system for using either a Sybase or an Oracle database. It also discusses
configuring a preexisting Sybase or Oracle database for use with Unicenter
AutoSys JM.
Note: You must complete the steps in this chapter prior to installing.
Since many of these preparatory steps require that you have UNIX superuser
permissions, you should log on to the server machine as “root.”
Installation Directory
All of the software, the database software and the data file (if you are installing a
bundled version) will be installed under the same directory. At this time, you
should confirm that you have sufficient available disk space, as described in
Disk Space in the chapter “Preparing for Installation,” in this guide.
You can specify any directory to be the installation directory, and it does not
have to be empty. By convention, this guide refers to the installation directory as
AUTOTREE, which represents the following default path:
/opt/CA/UnicenterAutoSysJM
After you complete the installation procedure, the following directories will exist
under AUTOTREE:
■ autosys
■ autouser
A UNIX user will be assigned ownership of all the files. This can be any user
except root.
If the user who is intended to own all files does not exist, create the user now.
This user’s home directory can be any directory except one of those found under
AUTOTREE.
Note: This user is not the only user that can run Unicenter AutoSys JM. Any
user with the correct environment can use Unicenter AutoSys JM.
If you are running NIS or NIS+, this step is required. If you are not running NIS
or NIS+, the internet demon is configured automatically when the installation
script is run; therefore, you can skip this step.
Unicenter AutoSys JM starts jobs by first connecting to the internet demon on the
client machine where the job is to be run. If your machine is running NIS or
NIS+, you must edit the services file, remake the map, and push it out to all
clients, including the server. Generally, the system administrator manages this
task.
where:
auto_remote Indicates the default service name for the remote agent. If multiple versions are
to be run concurrently on the same hardware, the remote agent service name
must be different for each version. In any case, the remote agent service name
must be the same in both the /etc/services file and the internet demon
configuration file (/etc/inetd.conf). For more information, see Modifying
Remote Agent Settings in the chapter “Configuring” of the Unicenter AutoSys
Job Management for UNIX User Guide.
5280 Indicates the default port number for the remote agent. If this number is
already in use by another internet service at your site, or if you wish to have
multiple versions of the remote agent installed on the same hardware, you can
change this number. In either case, this number should match the port number
recorded in both your server and client checklists.
After the /etc/services file is modified, it must be remade and pushed out to all
clients. For NIS, you would typically issue the following commands:
cd /var/yp
make
To see an updated services file from the client’s perspective, log on to the client
and issue the following command:
ypcat services
If you are using NIS+, see your system administrator for information on how to
push out the services file.
Note: These steps must be done successfully. The primary cause of a job not
starting on a client machine is the improper completion of these steps. If a client
machine that was working suddenly stops, it is usually because the
/etc/services file was modified.
The name of the bundled Sybase data server is AUTOSYSDB. The name of the
database in that data server is “autosys.”
Note: Bundled Sybase database users cannot change the size of the database that
is shipped with the software.
Sybase requires that changes be made to certain operating systems for the
database to work properly. The required changes, as dictated in the SYBASE
Adaptive Server Enterprise Installation Guide, are described following, by
operating system. The system administrator should do these changes. For more
information the SYBASE Adaptive Server Enterprise Installation Guide can be
obtained at supportconnect.ca.com.
Tru64
The server machine that will run bundled Sybase must be configured.
1. Edit the file named /etc/sysconfigtab and add the following line:
ipc:
<tab>shm-max = 96468992
where:
Note: If the above step is not performed, the server will not start.
HP-UX
Shared Memory Bundled Sybase will not start unless the operating system kernel has enough
shared memory available.
To adjust the shared memory value of the operating system, use the System
Administration Manager (SAM). Set the shared memory value to 67108864.
Note: Before executing the following instructions, shut down Adaptive Server
(or SQL Server).
1. From the SAM Kernel Configuration menu, choose Drivers and set the
Pending State for asyncdisk to In
2. From the Actions menu, rebuild the kernel, and restart the system.
3. At the UNIX prompt, execute the following statements as “root”. The user ID
of the user who is booting Adaptive Server and Backup Server must be the
owner of the /dev/async directory.
/etc/mknod /dev/async c 101 4
chmod 0660 /dev/async
chown sybase /dev/async
IBM AIX
To install bundled Sybase, you must enable asynchronous disk I/O regardless of
whether you intend to use it. You enable asynchronous I/O by adjusting various
kernel parameters using the SMIT utility. Setting the “servers” and “REQUEST”
parameters optimizes asynchronous I/O for the SQL Server. To address this
requirement, perform the following steps; you must be “root” to do the
following.
Enable asynchronous disk I/O by adjusting the kernel parameters, using the
System Management Interface Tool (SMIT).
If your system uses more than seven disks at the same time for
Asynchronous I/O, increase the MAXIMUM number of servers value by 1
for every active device.
Solaris 2.x
The Sybase SQL Server will not start unless the operating system kernel has
enough shared memory available. The operating system’s shared memory
parameter should always be set to a value larger than the value of the Sybase
SQL Server’s memory parameter.
To address this requirement, log on as “root” and perform the following steps:
If you want to use an existing Sybase database with Unicenter AutoSys JM, it
must be configured before the new software is installed. The Sybase system
administrator permission and the UNIX system administrator permissions are
required to perform these tasks.
The following checklist excerpted from the server checklist. It lists the decisions
you need to make when using an unbundled Sybase database with an Unicenter
AutoSys JM server. Steps describing these decisions are provided after the
checklist. You can fill in your responses in the checklist provided following, or
make a copy of the blank checklist following:
2. Determine the name of the Sybase SQL Server (DSQUERY) in which the
database will be created. Your Sybase database administrator should supply
this information.
3. Determine the name of the database that will contain the Unicenter AutoSys
JM information. The name can be anything; however, by convention it is
“autosys.”
4. Note the path to the Sybase interfaces file (SYBASE) on the server machine.
5. Identify the database user who has been granted both the System
Administrator (SA) and System Security Officer (SSO) roles.
Note: For Sybase SQL Server System 10 or greater, the installation script
prompts you to enter a database user and password granted the SA and SSO
roles. Users running unbundled Sybase 11.9.2 must accept the default user
(sa), then enter the sa account password.
7. Determine the size of the database that will contain the Unicenter AutoSys
JM information. The size can be any value greater than 25 MB. We
recommend 50 MB or 100 MB for large applications.
8. Create the database on the desired Sybase SQL Server using the following
entries (consult the Sybase documentation for more detailed instructions):
isql -Usa -Psa_password -S$DSQUERY
where:
After the previous command has completed executing, enter the following:
1> create database db_name on disk1=db_size
2> log on disk2=logfile_size
3> go
where:
db_name Is the name of the database to be created; it will contain the Unicenter AutoSys
JM data. This name is usually “autosys.”
disk1 Is the disk device to be used for storing the database information.
db_size Is the size of the database (in MB). This is often 60 MB.
logfile_size Is the size of the transaction log file (in MB). This is typically 25 MB.
The actual database objects (for example, tables and stored procedures) will be
installed when the installation script is run. These objects must be installed for
Unicenter AutoSys JM to run.
The recon utility is provided to verify that your Oracle environment is properly
configured for Unicenter AutoSys JM. This is described in the topic Oracle Setup
Configuration Utility in the chapter “Server Installation for Oracle,” in this
guide.
The following checklist is excerpted from the server checklist. It lists the
decisions you need to make when using an existing Oracle database with an
Unicenter AutoSys JM server. Steps describing these decisions are provided after
the checklist. You can fill in your responses in the checklist below, or make a
copy of the blank checklist following:
2. Determine the TNS alias name of the data server in which the tablespace will
be created.
3. Confirm that the TNS alias name is configured and that a SQL* Net V2
connect descriptor is in the tnsnames.ora configuration file.
4. Determine the name of the tablespaces that will contain the Unicenter
AutoSys JM database objects. The names can be anything; however, by
convention they are:
data tablespace ‘autodata’
index tablespace ‘autoindexes’
5. Determine the size of the tablespaces that will contain the database objects.
The data tablespace must be a minimum of 100 MB, 140 MB is
recommended. The index tablespace must be a minimum of 40 MB, 50 MB is
recommended. These tablespace sizes can be increased for large systems;
however, they should be increased in increments that are proportional to
each other.
7. Determine the name of the database user who has been granted DBA role.
Typically this is SYSTEM (the Oracle system user), but it does not have to be.
8. Note the password of the database user who has been granted DBA role.
Note: For optimal performance, we strongly recommend that the data files
for the data tablespace and the index tablespace reside on different hard
disks.
The following example creates the data tablespace data files and the index
tablespace data file.
sqlplus system/system_password@TNS_alias
where:
The actual database objects (for example, tables and stored procedures) will
be installed when the installation scripts are run. These database objects
must be installed for Unicenter AutoSys JM to work properly.
Server Installation
The following is an example server checklist that has been filled in for a basic
configuration with a bundled Sybase database. The procedures in this chapter
will use the information in this example checklist to illustrate the various
installation steps.
Server Checklist
Item Description
Platform Sun
Host ID 1234567ab
Server Checklist
Item Description
Password of root
Server Checklist
Item Description
Password of database user who has
been granted both SA and SSO roles
Using the Setup Wizard to Extract Unicenter AutoSys JM from the Media
The version media contains executables for Sybase databases. The media also
includes a script for loading the correct files for your platform and a text file
named README that describes the contents of the media.
3. Mount your CD-ROM following the appropriate steps for your operating
system. For example:
mount -r -t cdfs /dev/rz4c /cdrom
Directories
Notes:
In the eTrust client (seosdb), you must specify an admin user that is defined in
the eTrust server (pmdb). If this is not done then client (seosdb) will not be
synchronized with the server (pmdb).
3. If you have not changed directories since loading the release, change to the
install directory by entering:
cd autosys/install
Note: For unbundled Sybase users, make sure the data server is up. For
bundled Sybase users, the database does not have to be up.
If you are using NIS or NIS+, ensure that the steps described in the topic NIS or
NIS+ Only: Configure Internet Demon Services in the chapter “Preparing the
Server for Installation”, in this guide, have been completed. The services file
must be edited before you can successfully run auto_install.
At various points during the installation, you will be asked to verify the
information that has been entered. When asked, you can enter n (no) to go back
and correct any mistakes. At these verification prompts, you can cancel the
installation and start over.
The script always asks for confirmation before any system files are modified.
After you provide all the information the script needs to begin the installation, a
message that begins with the following line is displayed:
All the information needed to install AutoSys has been collected
Do you wish to proceed? ([y]|n)>
Bundled Sybase
For bundled Sybase installations, you are not prompted for the system
administrator user and password. The default “sa” user, and “sysadmin”
password are used.
Unbundled Sybase
For unbundled Sybase installation, the install script asks for information
regarding the existing data server and for a database user and password.
Up to this point, you can cancel the installation and start over. If you press Enter,
files will be modified and you will not be able to stop the installation. You will
see the following message followed by a series of informational messages
describing the progress of the installation:
Continuing installation...
If you want to make changes to the installation settings after this point, you can
rerun the installation after completion (see the topic Rerunning the Installation in
this chapter.)
The auto_install script can be run again if necessary. For bundled installations,
rerunning the script gives you the opportunity to overwrite the internet demon
configuration that was made before the previous installation.
The following procedures bring up the data server and event processor on the
server machine. They also test the connection to the remote agent on the server
machine, to ensure that the installation is successful to this point.
Note: The DNS server must be started before starting Unicenter AutoSys JM.
Before you start the event processor, you must ensure that the TZ environment
variable is set. The event processor references this setting to determine the
default time zone. Jobs with time-based starting conditions that do not specify a
time zone will have their start event scheduled based on the time zone under
which the event processor is running. This time zone is also used to report event
times, using the autorep command.
For C shell:
setenv TZ timezone
where:
2. Set up the environment for this user by sourcing the proper file in
AUTOUSER. (Since AUTOUSER is not defined yet, you must enter the full
path.)
In the following examples, “fiji” is the name of the server machine and
/opt/CA/UnicenterAutoSysJM represents the path to the directory where you
installed Unicenter AutoSys JM.
If you installed Unicenter AutoSys JM with a bundled database, the auto install
program automatically starts the dataserver. However, if you reboot, or shut
down, the dataserver will be shutdown and you must restart the dataserver
using the following procedures. (For unbundled Sybase users, the data server is
up already.)
or:
$SYBASE/ASE-12_0/install/RUN_data server_name
As the data server comes up, messages will appear on your screen. These are
messages from the data server. If there are problems, messages will continue to
be output to this window. That is why we recommend that you start the data
server in the console window.
When the output display is complete, press Enter a few times; the command
prompt should reappear.
Before you start the event processor, you must ensure that the TZ environment
variable is set. For more information on the TZ environment, see the section
Time Zone Setting, in this guide.
The event processor references this setting to determine the default time zone.
When eventor is run, you will lose control of the open command prompt
window. To run any other commands while the eventor is running you must
open a new command window. This will let you see the eventor output while
running commands through the second command prompt.
Or:
Running eventor –q allows the eventor to run silently, and returns a command
prompt. This will let you manually scroll through the eventor output without
losing the command prompt.
Notes: The eventor script is designed to make sure the environment is in the
right state before starting the event processor; specifically, it ensures that another
event processor is not already running.
If you installed eTrust AC, you must set your subscriber security word, before
proceeding. For more information on how to set the subscriber security word,
see the section eTrust Access Control in the chapter Security of the Unicenter
AutoSys Job Management for UNIX User Guide.
If you selected not to install eTrust AC, you must establish your EDIT and EXEC
superusers. If you selected to install eTrust, skip this topic.
The exec superuser is the only user who has permission to stop the event
processor. The exec superuser can also start and stop all jobs, regardless of their
ownership or permissions.
The edit superuser is the only user who can change the database password and
remote authentication method, change the owner of a job, or edit any job,
regardless of who owns it.
Notes: Until the exec superuser is defined, the event processor cannot be
properly shut down with a sendevent -E STOP_DEMON command.
To define the exec and edit superusers, execute the autosys_secure program.
When autosys_secure is invoked, the following menu appears:
AutoSys Security Utility.
[1] Administer EDIT and EXEC superusers.
[2] Change AutoSys database password.
[3] Change AutoSys remote authentication method.
[4] Create AutoSys User@Host or Domain password.
[5] Change AutoSys User@Host or Domain password.
[6] Delete AutoSys User@Host or Domain password.
[7] Enable eTrust Access Control
[8] Exit autosys_secure
[A] Get Encrypted Password for Adapters
>
The first time option [1] in the autosys_secure menu is chosen after Unicenter
AutoSys JM is installed, you are given another menu for edit and exec
superusers. Select one of the displayed options to continue. Both these privileges
can be assigned to the same user.
Important! Use caution when assigning edit and exec superuser privileges.
These privileges allow a user to modify behavior and change jobs.
AutoSys EDIT Super-User: autosys
...changed.
AutoSys EXEC Super-User: autosys
...changed.
After the initial assignments, only the edit superuser can change it. Option [1]
displays the current settings and allows the edit superuser to accept the same
users by pressing Enter, or change the users by entering a new specification.
The autosys_secure command is also used to change the database password for
the “autosys” user (the default password is “autosys”) and to enable remote
authentication. These options are explained in autosys_secure in the chapter
“Commands” of the Unicenter AutoSys Job Management for UNIX and
Windows Reference Guide.
Note: Any user with the proper environment can use Unicenter AutoSys JM;
you do not have to log on as the “autosys” user to use Unicenter AutoSys JM.
■ Check the machine’s database connection. If you are running dual event
servers, autoping checks both databases.
Internet Demon
If you are running NIS or NIS+, this verification is required. If you are not
running NIS or NIS+, the inetd is configured automatically when the installation
script is run; therefore, you can skip this step.
For more information on configuring the inetd, see the chapter Preparing the
Server for Installation, in this guide.
To verify that the remote agent on the server machine is functional, run the
autoping command. Assuming that the machine’s name is “fiji,” you would
enter:
autoping -m fiji
If you do not get this message, the remote agent is not configured properly, and
as a result, Unicenter AutoSys JM cannot start jobs on that machine (even if it is
the same machine as the server).
The remote agent is only executed when a connection is made to a client (that is,
a job is started). Therefore, looking in the process table for the remote agent
executable most likely will not be very informative.
If the remote agent is not installed properly, you can still verify the environment
and install the license keys; however, you will not be able to run a job.
If you have any problems, see the topic Remote Agent Troubleshooting in the
chapter “Troubleshooting” in the Unicenter AutoSys Job Management for UNIX
User Guide.
If you do not get this message, the database is not accessible. For troubleshooting
information, see the chapter “Troubleshooting” in the Unicenter AutoSys Job
Management for UNIX User Guide.
If you have not received your keys, contact Computer Associates Technical
Service at supportconnect.com.
You should install your license keys during the configuration of your Unicenter
AutoSys JM environment. However, if you do not install your license keys, you
will still be able to run jobs during the 30 day trial period. During this trial
period and thereafter you will receive license key warning and error messages.
Set up Autotrack
Now would be a good time to set up your desired autotrack tracking level.
autotrack tracks changes to the database (for example, job definition changes,
sendevent calls, and job overrides) and writes this information to the database.
Changes to job definitions made via JIL or the GUI can be tracked. Changes
made directly to the database through SQL commands cannot be tracked.
When you query for this information, autotrack prints a report to the screen, or
you can use standard UNIX file redirection to save the output to a file.
■ Sites where multiple users have permission to edit job definitions or send
AutoSys events.
Tip: Double quotes are required around the to_time and from_time
arguments.
To start tracking, use the autotrack -u command to set the tracking level to 1 or 2,
depending on the amount of detail you want tracked. By default, autotrack is set
to level 0 (no tracking). Only the exec superuser or edit superuser can change the
tracking level. For a complete description of this command, see the chapter
“Commands” of the Unicenter AutoSys Job Management for UNIX and
Windows Reference Guide.
After you have installed the license keys, you will be ready to enter a test job to
verify that the database can be updated. A client key must be installed on any
machine on which you want to run jobs, including the server machine if you
want to run jobs on that machine.
Note: If the instance is being controlled by eTrust AC, then prior to executing
any command line interface or GUI programs ensure the Unicenter AutoSys JM
subscriber security word is set first through running the AutoSys Secure Utility.
A test job called “test_install” is included with this version. The job definition is
in the file named $AUTOSYS/test/jil/test_install. If you view this job definition
in a text editor, you will see the following:
# JIL file to test the installation
# It will write a line to the Output File
insert_job: test_install
machine: localhost
command: /bin/echo “AUTOSYS install test”
std_out_file: /tmp/test_install.out
std_err_file: /tmp/test_install.err
Job definitions are specified using Job Information Language (or JIL). The jil
command is a language processor that parses the language and updates the
database. JIL is documented in the Unicenter AutoSys Job Management for
UNIX User Guide.
If your machine is not aliased to localhost, you will have to modify the
“test_install” job to specify your machine’s actual name.
3. To insert the “test_install” job into the database, enter the following
command at the operating system prompt:
jil < $AUTOSYS/test/jil/test_install
If there is a problem, you will see the following message and some error
messages:
Database change was NOT successful
Exit Code = 1
To run the “test_install” job, an event must be sent to cause the job to start.
4. To send the event to start this job, enter the following (ksh or csh):
sj test_install
This command starts the job “test_install” by using the sj alias that is defined
in the environment file. The sj alias represents the full command line shown
following (which could also be used to start the job, if you do not have the
aliases defined):
sendevent -E STARTJOB -J test_install
The event to start the job is now in the database, but the job itself will not
start until the event processor is up and running.
To verify that the job started and ran successfully, monitor the event processor
output log with the following command.
autosyslog -e
If the job ran successfully, the following message will be written to the
/tmp/test_install.out file:
AUTOSYS install test
If the job did not run successfully, you should see an error message indicating a
problem in the /tmp/test_install.err file.
After verifying the database is up, testing the database connection, and running
a test job, you need to test navigation through the data server. If this is
accomplished successfully, it confirms that the environment variables needed for
Unicenter AutoSys JM have been set up properly.
If all these steps have completed successfully, you have a properly installed
instance of Unicenter AutoSys JM.
Before you begin the client installation, review the chapter “Preparing for
Installation” in this guide for information about the client configuration required
before Unicenter AutoSys JM can be installed. Also, ensure that the server was
installed successfully and you completed your client checklist. Use the
information from your checklist during the installation.
Note: If you will be using the server machine as a client also—that is, running
jobs and utilities on the server machine—you do not need to install the client
software on the server. The client software is installed during the server
installation.
Client Installation
Unicenter AutoSys JM can use a client machine in several ways. It can be used as
the following:
It is possible to install the client for running jobs only, but it is better to install it
for all of the uses, as described in this chapter. If you want a client machine to
only start jobs, you can select that option during the installation.
The following is an example client checklist that has been filled in for a basic
configuration with bundled Sybase. The information in this checklist will be
used to illustrate the various installation steps.
Client Checklist
Item Description
Platform Sun
Host ID 987654yz
Password of root
Client Checklist
Item Description
configuration and output files
(AUTOUSER)
Sybase Database
Remote Agent
Full pathname to remote agent /usr/local/bin
executable
For more information about CCI see the appendix “Introducing CCI,” in this
guide.
If you are running NIS or NIS+, this step is required. If you are not running NIS
or NIS+, you can skip this step because the internet demon configuration will be
done for you automatically when the install scripts run.
Unicenter AutoSys JM starts jobs by first connecting to the internet demon on the
client machine where the job is to run. However, if the client machine is running
NIS or NIS+, you must edit the services file, remake the map, and push it out to
all clients, including the clients. Generally, the system administrator manages
this task and should be consulted.
See your system administrator for information on how to push out the services
file when running NIS+. For NIS systems, use the following procedures.
1. To change the services file, add the following line to /etc/services on the
server machine:
auto_remote 5280/tcp
The default port number for the remote agent is 5280. If this number is
already in use by another internet service at your site, or if you wish to have
multiple versions of the remote agent installed on the same hardware, you
can change this number. In either case, this number should match the port
number recorded in both your server and client checklists.
Note: The default service name for the remote agent is auto_remote. If
multiple versions are to be run concurrently on the same hardware, the
remote agent service names must be different for each version. In any case,
the remote agent service name must be the same in both the /etc/services
file and the internet demon configuration file (/etc/inetd.conf). For
information, see the chapter “Advanced Configurations,” in this guide.
After the /etc/services file has been modified, it must be remade and
pushed out to all clients. To do this for NIS, you would issue commands
similar to the following:
cd /var/yp
make
To see an updated services file from the client’s perspective, log on to the
client and issue the following command:
ypcat services
Note: It is important that these steps are done carefully and successfully. The
primary cause of a job not starting on a client machine is the improper
completion of the steps shown previously. If a client machine that was working
suddenly stops, it is usually because the /etc/services file was modified.
3. Mount your CD-ROM following the appropriate steps for your operating
system. For example:
mount -r -t cdfs /dev/rz4c /cdrom
AUTOTREE/autosys/install
This script creates required directories, configures special files for your
environment, installs runtime procedures, and (for networks that are not NIS or
NIS+) configures the Internet demon.
3. The installation must be done by “root.” If you have not already done so, log
on as “root.”
At various points during the installation, you will be asked to verify the
information that has been entered. When asked, you can enter n (no) to go back
and correct any mistakes. At these verification prompts, you can cancel the
installation and start over. The script always asks for confirmation before any
system files are modified.
After the script has been provided with all the information it needs to begin the
installation, a message beginning with the following line is displayed:
All the information needed to install AutoSys has been collected...
Do you wish to proceed? ([y]|n)>
Important! If you confirm at this point, you cannot cancel the installation.
Bundled Sybase
The following option allows you to select either all clients, or jobs only. If you
select “Jobs only,” the machine will be configured to execute jobs only—no
utilities can be used and the other client software (for example, the GUIs) will be
removed at the end of the installation.
Choose the type of client installation.
(1) All clients.
(2) Jobs only.
[1]> Enter
When installing with bundled Sybase, the interfaces file will be created for you.
For more information about CCI, see the appendix “Introducing CCI,” in this
guide.
After installing, the installation script continues with remote agent information,
described in the topic, Remote Agent Installation section in this chapter.
Unbundled Sybase
The following option allows you to select either all clients, or jobs only. If you
select “Jobs only,” the machine will be configured to execute jobs only—no
utilities can be used and the other client software (for example, the GUIs) will be
removed at the end of the installation.
Choose the type of client installation.
(1) All clients.
(2) Jobs only.
[1]> Enter
When installing with an unbundled Sybase data server, the interfaces file must
already exist on the client machine.
For more information about CCI, see the appendix “Introducing CCI,” in this
guide.
After the summary, the installation script continues with remote agent
information, described in the topic, Remote Agent Installation in this chapter.
Up to this point, you can cancel the installation and start over. If you press Enter,
files will be modified and you will not be able to stop the installation.
The following two lines of output would appear if you had installed jobs only.
AutoSys client installation for jobs only. Removing additional AutoSys clients...
AutoSys installation is complete.
If you do not get this message, the remote agent is not configured properly; and
as a result, AutoSys cannot start jobs on that machine.
The remote agent is only executed when a connection is made to a client (that is,
a job is started). Therefore, looking in the process table for the remote agent
executable most likely will not be very informative.
If the remote agent is not installed properly, you can still verify the environment
and install the license keys; however, you will not be able to run a job.
If you have any problems, see the topic Remote Agent Troubleshooting in the
chapter “Troubleshooting” in the Unicenter AutoSys Job Management for UNIX
User Guide.
Unicenter AutoSys JM uses the internet demon to start up the remote agent on
the client machine. To manually set up a machine to use the internet demon
(inetd), you must perform the steps described following.
Set Up Services
If you are using NIS or NIS+, the services file must be updated on the server, the
map must be remade and pushed out to the client. Generally, the system
administrator manages this task. (For details, see NIS or NIS+ Only: Configure
the Internet Demon Services in this chapter.)
If you are using NIS or NIS+, confirm that the NIS/NIS+ Services are set up
properly, and proceed with Step 2 following.
If you are not using NIS, the services are generally located in the
/etc/services file.
The default port number for the remote agent is 5280. If this number is
already in use by another internet service at your site, or if you wish to have
multiple versions of the remote agent installed on the same hardware, you
can change this number. In either case, this number should match the port
number recorded in both your server and client checklists.
Note: The default service name for the remote agent is “auto_remote.” If
multiple versions are to be run concurrently on the same hardware, the
remote agent service names must be different for each version. In any case,
the remote agent service name must be the same in both the /etc/services
file and the internet demon configuration file (/etc/inetd.conf). For more
information, see the chapter “Advanced Configurations,” in this guide.
2. As “root,” edit the file named /etc/inetd.conf and add the following single
line:
auto_remote stream tcp nowait root
/opt/CA/UnicenterAutoSysJM/autosys/bin/auto_remote auto_remote
where:
Sometimes, a kill -1 is not sufficient to reset the inetd. If you cannot connect
to the remote agent, you may have to issue a kill -9, then restart inetd. If
necessary, check with your system administrator for assistance.
Sybase Interfaces
For Sybase, there must be an entry in the interfaces file that corresponds to the
data server (in our example, AUTOSYSDB) that will host the database. If this is
not done properly, the remote agent cannot send events back to Unicenter
AutoSys JM.
The #AUTOENV# string is a special environment descriptor just for the remote
agent, and SYBASE points to the actual directory where the interfaces file
resides. Do not remove this descriptor from the /etc/auto.profile file.
For more information about the format and contents of the interfaces file, see
Configuring AutoSys for Dual-Event Servers in the chapter “Advanced
Configurations,” in this guide.
Profile Script
While the script can be specified at the individual job level, a default script must
be installed on every machine. An example script can be found in the file
$AUTOSYS/install/data/auto.profile. The default script should be copied into
/etc/auto.profile.
Executables
Unicenter AutoSys JM uses a Motif graphical user interface for all user interface
screens and dialogs. The application defaults file defines the GUI display. The
installation script installs this file in /usr/lib/X11/app-defaults or
/usr/openwin/lib/app-defaults. Consult your system administrator to verify
which directory is used on your system.
To install the GUI files yourself, you must copy the appropriate files for your
type of monitor, based on the following:
Sun Workstations
For Sun Workstations running OPEN WINDOWS, the directories are different.
The following files should be installed in $OPENWINHOME/lib/app-defaults:
$AUTOSYS/install/data/Autosc.*
$AUTOSYS/install/data/Autocons.*
$AUTOSYS/install/data/Autocal.*
$AUTOSYS/install/data/Xpert.*
Use the following commands (In this example, we are using the files for a color
monitor.):
cp $AUTOSYS/install/data/Autosc.co $OPENWINHOME/lib/app-defaults/Autosc
cp $AUTOSYS/install/data/Autocons.co $OPENWINHOME/lib/app-defaults/Autocons
cp $AUTOSYS/install/data/Autocal.co $OPENWINHOME/lib/app-defaults/Autocal
cp $AUTOSYS/install/data/Xpert.co $OPENWINHOME/lib/app-defaults/Xpert
The GUI screens will run under the OPEN LOOK window manager; however, an
additional key-binding file must be installed for cursor control to work properly.
To do that, enter the following command:
cp $AUTOSYS/install/data/XKeysymDB /usr/lib/X11
For Sun Workstations running OPEN WINDOWS, you should also copy
XKeysymDB to $OPENWINHOME/lib as shown following:
cp $AUTOSYS/install/data/XKeysymDB $OPENWINHOME/lib
Server Installation
The following is an example server checklist has been filled in for a basic
configuration, with an Oracle database. The installation procedures in this
chapter will use the information in this checklist to illustrate the various
installation steps.
Server Checklist
Item Description
Platform Sun
Host ID 1234567ab
Password of root
Server Checklist
Item Description
OS configuration completed (y/n) Yes
Oracle Database
Value of ORACLE_HOME /var/opt/oracle
Remote Agent
Full path name to remote agent /opt/CA/UnicenterAutoSysJM/autosys
executable /bin
For more information about CCI see the appendix “Introducing CCI,” in this
guide.
Using the Setup Wizard to Extract Unicenter AutoSys JM from the Media
The version media contains executables for Oracle databases. The media also
includes a script for loading the correct files for your platform and a text file
named README that describes the contents of the media.
3. Mount your CD-ROM following the appropriate steps for your operating
system. For example:
mount -r -t cdfs /dev/rz4c /cdrom
Notes:
The installation appears as a graphical interface if Java 2 runtime is available
in the path with suitable output. Otherwise, the installer uses a menu-driven
terminal window.
In the eTrust client (seosdb), you must specify an admin user that is defined
in the eTrust server (pmdb). If this is not done then client (seosdb) will not
be synchronized with the server (pmdb).
After you have extracted the files from the version, run the Oracle confirmation
utility described in the following section to verify that your database
environment is correct for Unicenter AutoSys JM.
To install the version, run the installation script as described in the topic
Running the Install Script in this chapter.
The recon utility is useful even after you have installed Unicenter AutoSys JM to
easily obtain information about your Oracle environment. If you are having
problems running with Oracle, run recon to more quickly diagnose the problem.
The recon utility will not change any settings on your system. Instead, it prints to
the screen the current settings and states whether or not these settings
correspond to what Unicenter AutoSys JM expects. Running this utility before
you install gives you the opportunity to correct Oracle settings, if necessary, and
protects you from a failed installation.
The recon utility analyzes the current user’s Oracle environment and the current
Oracle TNS configuration. For each accessible specified database, recon performs
the following tasks:
■ Checks for system packages, such as stored processes and functions from
catproc.sql.
The output of recon helps to determine what corrective action, if any, should be
taken before proceeding with the installation.
Syntax
user/password Specifies which user name and password to check. Used for all database access.
If you do not enter a password, you will be asked for it.
TNS_alias_list A space delimited list of TNS alias names of the data servers containing the
tablespaces.
Example
The following example prints verbose information about the Oracle setup:
The output of recon shows the path to, or status of the following:
■ SQL*Plus binary
The recon utility displays the results of its attempt to log on and inspect the
databases. Specifically, it checks:
For free tablespace information, the output will resemble the following:
The following example shows the same request with non-verbose output:
recon system/manager AUTOSYSDB
3. If you have not changed directories since loading the version, change to the
install directory by entering the following:
cd autosys/install
The install script (auto_install) will ask you for information as it executes,
supplying defaults whenever possible. For the basic configuration, we
recommend that you accept all the defaults.
If you are using NIS or NIS+, ensure that the steps described in the NIS or NIS+
Only: Configure Internet Demon Services see the section of the chapter,
“Preparing the Server for Installation,” in this guide, have been completed. The
services file must be edited before you can successfully run auto_install.
Note: The AutoSys software is compiled for either Sybase or Oracle prior to
shipment; it is not configurable during installation. The auto_install script prints
out which version you are installing.
At various points during the installation, you will be asked to verify the
information that has been entered. When prompted, you can enter n (no) to go
back and correct any mistakes. At these verification prompts, you can cancel the
installation and start over.
The script always prompts for confirmation before any system files are modified.
After the script has been provided with all the information it needs to begin the
installation, a message beginning with the following line is displayed:
All the information needed to install AutoSys has been collected...
Do you wish to proceed? ([y]|n)>
Important! If you confirm at this point, you cannot cancel the installation.
After the summary, the installation script continues with remote agent
information, described in the section Remote Agent Information.
Up to this point, you can cancel the installation and start over. If you press Enter,
files will be modified and you will not be able to stop the installation. You will
see the following message followed by a series of informational messages
describing the progress of the installation:
Continuing installation...
If you want to make changes to the installation settings after this point, you can
rerun the installation after completion (see Rerunning the Installation in this
chapter).
WARNING! For Oracle server installations, rerunning auto_install will cause the
database objects to be rebuilt; as a result, it will delete all the data that has been
entered into the AutoSys database.
For more information about installing or rebuilding a database, see the chapter
“Advanced Configurations,” in this guide.
The following procedures bring up the data server and event processor on the
server machine. They also test the connection to the remote agent on the server
machine, to ensure that the installation is successful to this point.
Note: The DNS server must be started before starting Unicenter AutoSys JM.
Before you start the event processor, you must ensure that the TZ environment
variable is set. The event processor references this setting to determine the
default time zone. Jobs with time-based starting conditions that do not specify a
time zone will have their start event scheduled based on the time zone under
which the event processor is running. This time zone is also used to report event
times, using the autorep command.
For C shell:
setenv TZ timezone
where:
2. Set up the environment for this user by sourcing the proper file in
AUTOUSER. (Since AUTOUSER is not defined yet, you must enter the full
path.)
In the following examples, “fiji” is the name of the server machine, cd to your
autouser directory, and run the following command for your shell.
If you installed Unicenter AutoSys JM with a bundled database, the auto install
program automatically starts the dataserver. However, if you reboot, or shut
down, the dataserver will be shutdown and you must restart the dataserver
using the following procedures. (For unbundled Sybase users, the data server is
up already.)
or:
$SYBASE/ASE-12_0/install/RUN_data server_name
As the data server comes up, messages will appear on your screen. These are
messages from the data server. If there are problems, messages will continue to
be output to this window. That is why we recommend that you start the data
server in the console window.
When the output display is complete, press Enter a few times; the command
prompt should reappear.
Before you start the event processor, you must ensure that the TZ environment
variable is set. For more information on the TZ environment, see the section
Time Zone Setting, in this guide.
The event processor references this setting to determine the default time zone.
When eventor is run, you will lose control of the open command prompt
window. To run any other commands while the eventor is running you must
open a new command window. This will let you see the eventor output while
running commands through the second command prompt.
Or:
Running eventor –q allows the eventor to run silently, and returns a command
prompt. This will let you manually scroll through the eventor output without
losing the command prompt.
Notes: The eventor script is designed to make sure the environment is in the
right state before starting the event processor; specifically, it ensures that another
event processor is not already running.
If you installed eTrust AC, you must set your subscriber security word, before
proceeding. For more information on how to set the subscriber security word,
see the section eTrust Access Control in the chapter Security of the Unicenter
AutoSys Job Management for UNIX User Guide.
If you selected not to install eTrust AC, you must establish your EDIT and EXEC
superusers. If you selected to install eTrust, skip this topic.
The exec superuser is the only user who has permission to stop the event
processor. The exec superuser can also start and stop all jobs, regardless of their
ownership or permissions.
The edit superuser is the only user who can change the database password and
remote authentication method, change the owner of a job, or edit any job,
regardless of who owns it.
Notes: Until the exec superuser is defined, the event processor cannot be
properly shut down with a sendevent -E STOP_DEMON command.
To define the exec and edit superusers, execute the autosys_secure program.
When autosys_secure is invoked, the following menu appears:
AutoSys Security Utility.
[1] Administer EDIT and EXEC superusers.
[2] Change AutoSys database password.
[3] Change AutoSys remote authentication method.
[4] Create AutoSys User@Host or Domain password.
[5] Change AutoSys User@Host or Domain password.
[6] Delete AutoSys User@Host or Domain password.
[7] Enable eTrust Access Control
[8] Exit autosys_secure
[A] Get Encrypted Password for Adapters
>
The first time option [1] in the autosys_secure menu is chosen after Unicenter
AutoSys JM is installed, you are given another menu for edit and exec
superusers. Select one of the displayed options to continue. Both these privileges
can be assigned to the same user.
Important! Use caution when assigning edit and exec superuser privileges.
These privileges allow a user to modify behavior and change jobs.
AutoSys EDIT Super-User: autosys
...changed.
AutoSys EXEC Super-User: autosys
...changed.
After the initial assignments, only the edit superuser can change it. Option [1]
displays the current settings and allows the edit superuser to accept the same
users by pressing Enter, or change the users by entering a new specification.
The autosys_secure command is also used to change the database password for
the “autosys” user (the default password is “autosys”) and to enable remote
authentication. These options are explained in autosys_secure in the chapter
“Commands” of the Unicenter AutoSys Job Management for UNIX and
Windows Reference Guide.
Note: Any user with the proper environment can use Unicenter AutoSys JM;
you do not have to log on as the “autosys” user to use Unicenter AutoSys JM.
■ Check the machine’s database connection. If you are running dual event
servers, autoping checks both databases.
Internet Demon
If you are running NIS or NIS+, this verification is required. If you are not
running NIS or NIS+, the inetd is configured automatically when the installation
script is run; therefore, you can skip this step.
For more information on configuring the inetd, see the chapter Preparing the
Server for Installation, in this guide.
To verify that the remote agent on the server machine is functional, run the
autoping command. Assuming that the machine’s name is “fiji,” you would
enter:
autoping -m fiji
If you do not get this message, the remote agent is not configured properly, and
as a result, Unicenter AutoSys JM cannot start jobs on that machine (even if it is
the same machine as the server).
The remote agent is only executed when a connection is made to a client (that is,
a job is started). Therefore, looking in the process table for the remote agent
executable most likely will not be very informative.
If the remote agent is not installed properly, you can still verify the environment
and install the license keys; however, you will not be able to run a job.
If you have any problems, see the topic Remote Agent Troubleshooting in the
chapter “Troubleshooting” in the Unicenter AutoSys Job Management for UNIX
User Guide.
If you do not get this message, the database is not accessible. For troubleshooting
information, see the chapter “Troubleshooting” in the Unicenter AutoSys Job
Management for UNIX User Guide.
If you have not received your keys, contact Computer Associates Technical
Service at supportconnect.com.
You should install your license keys during the configuration of your Unicenter
AutoSys JM environment. However, if you do not install your license keys, you
will still be able to run jobs during the 30 day trial period. During this trial
period and thereafter you will receive license key warning and error messages.
Set up Autotrack
Now would be a good time to set up your desired autotrack tracking level.
autotrack tracks changes to the database (for example, job definition changes,
sendevent calls, and job overrides) and writes this information to the database.
Changes to job definitions made via JIL or the GUI can be tracked. Changes
made directly to the database through SQL commands cannot be tracked.
When you query for this information, autotrack prints a report to the screen, or
you can use standard UNIX file redirection to save the output to a file.
■ Sites where multiple users have permission to edit job definitions or send
AutoSys events.
Tip: Double quotes are required around the to_time and from_time
arguments.
To start tracking, use the autotrack -u command to set the tracking level to 1 or 2,
depending on the amount of detail you want tracked. By default, autotrack is set
to level 0 (no tracking). Only the exec superuser or edit superuser can change the
tracking level. For a complete description of this command, see the chapter
“Commands” of the Unicenter AutoSys Job Management for UNIX and
Windows Reference Guide.
After you have installed the license keys, you will be ready to enter a test job to
verify that the database can be updated. A client key must be installed on any
machine on which you want to run jobs, including the server machine if you
want to run jobs on that machine.
Note: If the instance is being controlled by eTrust AC, then prior to executing
any command line interface or GUI programs ensure the Unicenter AutoSys JM
subscriber security word is set first through running the AutoSys Secure Utility.
A test job called “test_install” is included with this version. The job definition is
in the file named $AUTOSYS/test/jil/test_install. If you view this job definition
in a text editor, you will see the following:
# JIL file to test the installation
# It will write a line to the Output File
insert_job: test_install
machine: localhost
command: /bin/echo “AUTOSYS install test”
std_out_file: /tmp/test_install.out
std_err_file: /tmp/test_install.err
Job definitions are specified using Job Information Language (or JIL). The jil
command is a language processor that parses the language and updates the
database. JIL is documented in the Unicenter AutoSys Job Management for
UNIX User Guide.
If your machine is not aliased to localhost, you will have to modify the
“test_install” job to specify your machine’s actual name.
3. To insert the “test_install” job into the database, enter the following
command at the operating system prompt:
jil < $AUTOSYS/test/jil/test_install
If there is a problem, you will see the following message and some error
messages:
Database change was NOT successful
Exit Code = 1
To run the “test_install” job, an event must be sent to cause the job to start.
4. To send the event to start this job, enter the following (ksh or csh):
sj test_install
This command starts the job “test_install” by using the sj alias that is defined
in the environment file. The sj alias represents the full command line shown
following (which could also be used to start the job, if you do not have the
aliases defined):
sendevent -E STARTJOB -J test_install
The event to start the job is now in the database, but the job itself will not
start until the event processor is up and running.
To verify that the job started and ran successfully, monitor the event processor
output log with the following command.
autosyslog -e
If the job ran successfully, the following message will be written to the
/tmp/test_install.out file:
AUTOSYS install test
If the job did not run successfully, you should see an error message indicating a
problem in the /tmp/test_install.err file.
After verifying the database is up, testing the database connection, and running
a test job, you need to test navigation through the data server. If this is
accomplished successfully, it confirms that the environment variables needed for
Unicenter AutoSys JM have been set up properly.
If all these steps have completed successfully, you have a properly installed
instance of Unicenter AutoSys JM.
Before you begin the client installation, review the chapter “Preparing for
Installation,” in this guide, which describes the client configuration required
before Unicenter AutoSys JM can be installed. Also, ensure that the server has
been successfully installed and your client checklist has been filled out. Use the
information in your checklist during the installation.
Note: If you will be using the server machine as a client that is running jobs and
utilities on the server machine, you do not need to install the client software on
the server. The client software is installed during the server installation.
Client Installation
Unicneter AutoSys JM can use a client machine in several ways. It can be used
as:
It is possible to install the client for running jobs only, but it is better to install it
for all of the above uses, as described in this chapter. If you want a client
machine to only start jobs, you can select that option during the installation.
Following is an example client checklist that has been filled in for a basic
configuration with Oracle. The information in this checklist will be used to
illustrate the various installation steps.
Client Checklist
Item Description
Platform Sun
Host ID 987654yz
Password of root
Client Checklist
Item Description
(AUTOUSER)
Oracle Database
Value of ORACLE_HOME /var/opt/oracle
Remote Agent
Client Checklist
Item Description
Port number for the remote agent 5280
For more information on CCI, see the appendix “Introducing CCI,” in this guide.
These steps assume the installation directory into which the Unicneter AutoSys
JM directory (AUTOSYS) is to be copied is;
/usr/vendor/autotree
3. Mount your CD-ROM following the appropriate steps for your operating
system. For example:
mount -r -t cdfs /dev/rz4c /cdrom
After you have extracted the files from the version, run the Oracle confirmation
utility described in the following section to verify that your database
environment is correct for Unicenter AutoSys JM.
To install the version, run the installation script as described in the topic
Running the Install Script in this chapter.
3. The installation must be done by “root”; if you have not already done so, log
on as “root.”
At various points during the installation, you will be asked to verify the
information that has been entered. When prompted, you can enter n (no) to go
back and correct any mistakes. At these verification prompts, you can cancel the
installation and start over.
The script prompts for confirmation before any system files are modified.
After you provide all the information the script needs to begin the installation, a
message that begins with the following line is displayed:
All the information needed to install AutoSys has been collected...
Do you wish to proceed? ([y]|n)>
Important! If you confirm at this point, you cannot cancel the installation.
The following option lets you select either all clients, or jobs only. If you select
“Jobs only,” the machine will be configured to execute Unicneter AutoSys JM
jobs only—no Unicneter AutoSys JM utilities can be used and the other client
software (for example, the GUIs) will be removed at the end of the installation.
Choose the type of client installation.
(1) All clients.
(2) Jobs only.
[1]> Enter
After the summary, the installation script continues with remote agent
information, described in the topic “Remote Agent Verification,” in this chapter.
Up to this point, you can cancel the installation and start over. If you press Enter,
files will be modified and you will not be able to stop the installation. You will
see the following message:
Continuing installation...
The next two lines of output would appear if you had installed jobs only.
AutoSys client installation for jobs only. Removing additional AutoSys clients...
AutoSys installation is complete.
If you do not get this message, the remote agent is not configured properly, and
as a result, Unicneter AutoSys JM cannot start jobs on that machine.
The remote agent is only executed when a connection is made to a client (that is,
a job is started). Therefore, looking in the process table for the remote agent
executable most likely will not be very informative.
If the remote agent is not installed properly, you can still verify the environment
and install the license keys; however, you will not be able to run a job.
If you have any problems, see the Remote Agent Troubleshooting section in the
chapter “Troubleshooting” in the Unicenter AutoSys Job Management for UNIX
User Guide.
Unicneter AutoSys JM uses the internet demon to start up the remote agent on
the client machine. To manually set up a machine to use the demon (inetd), you
must perform the steps following.
Set Up Services
If you are using NIS or NIS+, the services file must be updated on the server, the
map must be remade and pushed out to the client by the system administrator.
For details see the topic NIS or NIS+ Only: Configure the Internet Demon
Services in this chapter.
If you are using NIS or NIS+, confirm that the NIS/NIS+ services are set up
correctly, and proceed to Step 2.
Note: If you are not using NIS, the services are located in:
/etc/services file
To set up services:
The default port number for the remote agent is 5280. If this number is
already in use by another internet service, or if you want to have multiple
versions of the remote agent installed on the same hardware, change this
number. In either case this number should match the port number recorded
in both your server and client checklist.
Note: The default service name for the remote agent is auto_remote. If
multiple versions of Unicneter AutoSys JM are to be run concurrently on the
same hardware, the remote agent service names must be different for each
version. Also, the remote agent service name must be the same in both the
/etc/services file and the internet demon configuration file /etc/inetd.conf.
For more information see the chapter “Advanced Configurations,” in this guide.
2. As “root,” edit the file named /etc/inetd.conf and add the following line:
Auto_remote stream tcp nowait root
/opt/CA/UnicenterAutoSysJM/sutosys/bin/auto_remote auto_remote
where:
At times a kill –1 is not sufficient to reset the inetd. If you are unable to
connect to the remote agent, you may have to issue a kill –9, then restart
inetd. Check with your system administrator for assistance.
Profile Script
While the script can be specified at the individual job level, a default script must
be installed on every machine. An example script can be found in the file
$AUTOSYS/install/data/auto.profile. The default script should be copied into
/etc/auto.profile.
Executables
where:
#AUTOENV# Indicates a special environment descriptor just for the remote agent.
Unicneter AutoSys JM uses a Motif graphical user interface for all user interface
screens and dialogs. The GUI display is defined by the application defaults file.
The installation script installs the application defaults file in:
/usr/lib/X11/app-defaults or /usr/openwin/lib/app-defaults
To install the GUI files yourself, you must copy the appropriate files for your
type of monitor:
Sun Workstations
For Sun Workstations running OPEN WINDOWS, the directories are different.
The following files should be installed in $OPENWINHOME/lib/app-defaults:
$AUTOSYS/install/data/Autosc.*
$AUTOSYS/install/data/Autocons.*
$AUTOSYS/install/data/Autocal.*
$AUTOSYS/install/data/Xpert.*
The GUI screens will run under the OPEN LOOK window manager; however, an
additional key binding file must be installed for cursor control to work properly.
To do that, enter the following:
cp $AUTOSYS/install/data/XKeysymDB /usr/lib/X11
For Sun Workstations running OPEN WINDOWS, you should also copy
XKeysymDB to $OPENWINHOME/lib.
Advanced Configurations
8
This chapter describes how to install and configure advanced features of
Unicenter AutoSys JM, including installing and configuring high-availability
options and configuring multiple instances for cross-instance job dependencies.
It describes only the configuration file parameters that are used to set these
advanced features. For complete information on the configuration, including a
sample file, see the chapter “Configuring” in the Unicenter AutoSys Job
Management for UNIX User Guide.
The $AUTOSERV value is the name of the instance with which the configuration
file is associated. Upon startup, Unicenter AutoSys JM reads the configuration
file to determine its behavior, including which databases to connect to and how
to react to certain error conditions. Also, the runtime behavior of the event
processor is based on the parameters found in this configuration file.
PORT NUMBER
The port on which Unicenter AutoSys Remote Command Service will listen
for remote command requests
EPLOG LINES
Number of lines from the Event Processor log to retrieve and return to the
web server when handling an EPLOG request
Note: After installation the PORT NUMBER and EPLOG LINES values cannot
be modified.
The installation allows the user to configure the VALID HOSTS and IPs and the
VALID DIRECTORIES files for the RCS. In order for the Unicenter AutoSys JM
Web Interface and the RCS to communicate, the hostname for the Unicenter
AutoSys JM Web Interface machine must be entered. In order to view the output
or error file through the Unicenter AutoSys JM Web Interface, the directory path
of the file must be entered.
After installation, the VALID HOST and IPs file and the VALID DIRECTORIES
file can be modified by editing the validips and valid_dirs files.
VALID DIRS
The valid_dirs file contains a list of valid directories that the RCS uses to find
an output or error file from a job that has run successfully. To view a job log
through the Web Interface, the directory path of the output or error file must
be specified in this file.
The valid directory refers to the directory that contains the standard output
or error file that was specified in the job definition.
For a user to view the output from the preceding job using the Web
Interface, the directory, /tmp must be specified as a valid directory. When
trying to view a log in which the directory is not an entry in the valid_dirs
file, the request will fail the Web Interface will display a message saying the
directory is not valid.
Note: RCS must be restarted after any modifications are made to either one
of these two files.
RCS is started and stopped through the start_rcs and stop_rcs scripts located in
the autosys/bin directory.
Logging
All RCS messages including Unicenter AutoSys JM Web Interface requests and
errors are written to the RCS logfile, asrcslsn.log. These messages are useful for
viewing when a job may have been added, modified or deleted from the Event
Server, when a web user mapping is performed, or an EP or Joblog request is
issued through the Unicenter AutoSys JM Web Interface. This logfile can also be
used for troubleshooting. For example; a user trying to add a job through an
unknown Unicenter AutoSys JM Web Interface (web interface machine is not
listed as a valid machine in the validips configuration file). In this example, the
RCS log will inform the user which Unicenter AutoSys JM Web Interface
machine must be added to the validips configuration file.
Daily at midnight, the RCS logfile rolls over and the previous day's file is
renamed to asrcslsn.previous_date. A new logfile is then created to receive the
new (current) day’s messages. Since there is a unique RCS logfile for each day, it
is suggested regular maintenance be performed on the logfiles to ensure the
availability of disk space.
/opt/CA/SharedComponents/UnicenterAutoSys/logfiles
Debugging
When a problem occurs, RCS can write more specific system debug information
to the logfile. This is controlled through the system environment variable:
RCSDEBUG
To have RCS write this information to the log file, the value must be set to 1.
Set and export the variable using the syntax of your current shell.
When processing events, the event processor reads from both event servers. If it
detects an event on one server and not on the other, it will copy the missing
event to the other server. In addition, the remote agent writes events to both
event servers.
To run in dual-server mode, you must add a second event server (database). The
event servers should run on different machines to prevent a single point of
failure.
If you are using bundled Sybase databases, you can install two databases and
modify the appropriate configuration parameters. If you are using unbundled
databases, you need to install and configure the two databases before you can
use them. Then, you can set the appropriate configuration parameters.
■ With dual-server mode, both databases must be of the same type (that is,
both Sybase or both Oracle).
■ The two event servers should reside on two different database servers,
running on different machines, to avoid a single point of failure.
■ The event processor will not start unless it can connect to both databases.
■ For bundled Sybase, you can install two servers following the
instructions in the chapter “Server Installation for Sybase,” in this guide.
If you are currently running in single server mode, copy your installed
bundled data server to a different machine and change its name, as
described in Copying Bundled Sybase following.
These steps are explained in this chapter for either a Sybase or an Oracle
database.
Bundled Sybase users who want to move from single-server to dual-server mode
should copy their installed bundled data server to a different machine and
change its name.
If the two machines have the same operating system, copy the software as
explained following.
Notes:
■ We recommend that the Sybase directory be local to the machine where
the second event server is to be run.
■ If the server installation directory is not visible on the second data server
machine, you need to log on to the second data server and perform two
remote copies. In the example following, fiji is the host name of the first
data server machine (the second command copies the auto_db_rename
script):
rcp -r fiji:/usr/vendor/autotree/sadb/ /usr/vendor/autotree/sadb2
rcp -r fiji:/usr/vendor/autotree/autosys/
install/usr/vendor/autotree/autosys/install
If your system does not support the rcp command, use another command
such as cp -r or tar.
3. Set the SYBASE environment variable to the location of the new data server,
like:
setenv SYBASE /usr/vendor/autotree/sadb2
or:
SYBASE=/usr/vendor/autotree/sadb2; export SYBASE
4. Change the name of the second data server from AUTOSYSDB to a different
name, such as AUTOSYSDB2, using the auto_db_rename script, like:
cd $AUTOSYS/install
./auto_db_rename
If the two machines do not have the same operating system, you must load the
data server from the media. The version media contains executables complied for
all supported platforms for the database you requested (Sybase or Oracle). The
media also includes a script for loading the correct files for your platform and a
text file named README that describes the contents of the media.
3. Change to the directory in which you will be installing the software, like:
cd /usr/vendor/autotree
For CD-ROM, mount your CD-ROM following the appropriate steps for
your operating system. For example:
mount –r –t cdfs /dev/rz4c /cdrom
Next, copy as_load.sh from the cdrom to the location where you want to
install Unicenter AutoSys JM. If as_load.sh is not executable, change its
permissions and run as_load from the cdrom. Usage:
./as_load.sh cdrom directory
5. The load script presents a numbered list of supported platforms. Enter the
number corresponding to the platform on which you want to install
Unicenter AutoSys JM. The load script extracts the appropriate files for the
chosen platform.
6. Set the SYBASE environment variable to the location of the new data server,
like:
setenv SYBASE /usr/vendor/autotree/sadb2
or:
SYBASE=/usr/vendor/autotree/sadb2; export SYBASE
7. Change the name of the second data server, for example, from AUTOSYSDB
to AUTOSYSDB2, using the auto_db_rename script, like:
cd $AUTOSYS/install
./auto_db_rename
To install a second unbundled Sybase event server on a machine that does not
have Unicenter AutoSys JM installed on it:
To install a second unbundled Sybase event server on a machine that already has
Unicenter AutoSys JM installed:
2. If you have not already done so, source the appropriate environment
variables. For details, see Environment in the chapter “Introduction,” in this
guide.
b. Into a new, empty database, load the database objects by issuing the
following command:
./auto_dbobj data_srvr sa_usr sa_pass db_name E
where:
sa_pass Indicates the database system administrator’s password. If you do not know
this password, check with your database administrator.
Installing Oracle
To install a second Oracle event server on a machine that does not have
Unicenter AutoSys JM installed on it:
■ The names and sizes of the tablespaces are the same as the first database.
To install a second Oracle event server on a machine that already has Unicenter
AutoSys JM installed:
2. If you have not already done so, source the appropriate environment
variables. (For details, see the section Environment in the chapter
“Introduction,” in this guide.)
where:
Data_Tablespace Indicates the name of the data tablespace. The default value is AUTODATA.
Index_Tablespace Indicates the name of the index tablespace. The default value is
AUTOINDEXES.
Temp_Tablespace Indicates the name of the temporary tablespace. The default value is TEMP.
Configuring Sybase
1. Modify the configuration file so that it points to two databases. Edit the
$AUTOUSER/config.$AUTOSERV file using any text editor.
In single server mode, there is only one event server listed in the
configuration file. A typical Sybase single server mode setting is:
# Specify the Databases
#
EventServer=AUTOSYSDB:autosys
# Sybase database example
#EventServer=SYBSERVER2:autosys
For example, assume you are working with the following setup:
In this scenario, you would modify the configuration file to add a line to
define the second event server, like:
EventServer=AUTOSYSDB:autosys
EventServer=AUTOSYSDB2:autosys
# Sybase database example
#EventServer=SYBSERVER2:autosys
There must not be a comment character (#) at the beginning of the line that
defines the second event server.
Place a comment character (#) at the beginning of the line used for single
server mode, like:
# DBEventReconnect=50
Remove the comment character (#) from the beginning of the line used for
Dual-Server Mode, like:
DBEventReconnect=50,5
The default setting for dual-server mode specifies that the event processor
should attempt five connections with the event servers. If after five times it
cannot connect, it should rollover to single server mode, marking the other
event server as “down.” Once in single server mode, the event processor
should attempt a connection fifty times, and if it is unsuccessful, the event
processor shuts down.
Similarly, upon start up, the event processor makes five attempts to connect
to the event servers, and if unsuccessful, it will immediately rollover to
single server mode, using the event server that is still functioning.
4. Next, you need to modify the interfaces files on all server and client
machines so that it has entries for both event servers.
On some platforms such as Solaris, the format of the interfaces file may look
like this, with entries for data servers residing on machines named fiji and
bermuda:
#
# AUTOSYSDB on fiji (192.9.100.90) using tcp
# services: query (6324) master (6324)
# console (6325)
#
AUTOSYSDB
query tli tcp /dev/tcp x:00>218b4c009645a
master tli tcp /dev/tcp x:00>218b4c009645a
console tli tcp /dev/tcp x:00>218b5c009645a
#
#
# AUTOSYSDB2 on bermuda (192.9.100.91) using tcp
# services: query (6324) master (6324)
# console (6325)
#
AUTOSYSDB2
query tli tcp /dev/tcp x:00>218b4c009645b
master tli tcp /dev/tcp x:00>218b4c009645b
console tli tcp /dev/tcp x:00>218b5c009645b
On other platforms, a typical interfaces file looks like the following example.
This example is for an AIX machine:
AUTOSYSDB
query tcp sun-ether fiji 6324
master tcp sun-ether fiji 6324
console tcp sun-ether fiji 6325
AUTOSYSDB2
query tcp sun-ether bermuda 6324
master tcp sun-ether bermuda 6324
console tcp sun-ether bermuda 6325
Note: In the interfaces file, a single tab must precede the first word of every line;
do not use spaces. A single space is used to delimit each element in an entry line.
Incorrect formatting will prevent communication with the database.
If the client and server machines have different operating systems and different
interfaces file formats, they cannot be directly copied from one to another. To
update the interfaces files on Solaris, Sequent, and NCR machines, use the autotli
command.
The autotli command creates an interfaces file data server entry and writes the
output to standard output. As a result, you can redirect this output to create
either a new interfaces file, or append an existing one.
The command following will add the AUTOSYSDB2 data server entry for
machine named “bermuda” and append it to an existing interfaces file. It can be
executed using C shell or Korn shell (the AUTOSYS variable must be set
already):
$AUTOSYS/install/autotli -s AUTOSYSDB2 -h bermuda - p6324 >> $SYBASE/interfaces
where:
>> Appends the information to the $SYBASE/interfaces file. (Be sure to use two
redirect symbols ( >> ); a single redirect ( > ) symbol will overwrite the existing
file.)
Configuring Oracle
1. Modify the configuration file so that it points to two databases. Edit the
$AUTOUSER/config.$AUTOSERV file using any text editor.
In single server mode, there is only one event server listed in the
configuration file. A typical Oracle single server mode setting is:
# Specify the Database(s)
#
EventServer=AUTOSYSDB
# EventServer=AUTOSYSDB:autosys
# Sybase database example
#EventServer=SYBSERVER2:autosys
# Oracle database example
#EventServer=ORASERVER2
For example, assume you are working with the following setup:
In this scenario, you would modify the configuration file to add a line to
define the second event server, like this (the added line is in bold font):
EventServer=AUTOSYSDB
EventServer=AUTOSYSDB2
There must not be a comment character (#) at the beginning of the line that
defines the second event server.
Place a comment character (#) at the beginning of the line used for single-
server mode, like:
# DBEventReconnect=50
Remove the comment character (#) from the beginning of the line used for
dual-server mode, like:
DBEventReconnect=50,5
The default setting for dual server mode specifies that the event processor
should attempt five connections with the event servers. If a connection is not
successful after five attempts, the event processor should rollover to single
server mode, marking the other event server as “down.” Once in single
server mode, the event processor should attempt a connection 50 times, and
if it is unsuccessful, the event processor shuts down.
Similarly, upon start up, the event processor makes five attempts to connect
to the event server, and if unsuccessful, it will immediately rollover to single
server mode, using the event server that is still functioning.
The second event server must be synchronized with the first server before you
can begin processing using dual-server mode. This section describes this process.
These instructions apply to both Sybase and Oracle.
Before you synchronize the databases, make sure that no one is creating or
editing job definitions (either through jil or the GUI dialogs), then stop the event
processor, as shown in the following command (you must be the exec superuser
to do this):
sendevent -E STOP_DEMON
where:
source_server Indicates the name of the source Sybase data server or Oracle instance.
target_server Indicates the name of the target Sybase data server or Oracle instance.
autosys_password Indicates the “autosys” user password. By default, this is “autosys,” but you
could have changed this password using the autosys_secure command.
source_db_name Indicates the name of the source Sybase or Oracle database. If omitted, this
defaults to “autosys.”
target_db_name Indicates the name of the target Sybase or Oracle database. If omitted, this
defaults to “autosys.”
For Sybase: The SYBASE environment variable must be set and the
$SYBASE/interfaces file must contain entries for both the source
(AUTOSYSDB) and the target (AUTOSYSDB2) data servers.
For Oracle: The ORACLE_HOME environment variable must be set and the
TNS names file must contain valid entries for both the source and target data
servers.
■ There is sufficient disk space for the autobcp temporary output file.
■ No one is executing JIL or using the GUI dialogs to change job definitions.
For this example, assume that a new, target data server named AUTOSYSDB2 is
to be synchronized with an existing, source data server named AUTOSYSDB. For
your own process, substitute your own data server names. This example also
assumes the source and target database names are the default “autosys,” and
therefore do not have to be specified in the command.
$AUTOSYS/install/autobcp AUTOSYSDB AUTOSYSDB2 \
/tmp/dumpfile autosys | tee /tmp/autobcp.out
where:
AUTOSYSDB Indicates the name of the source Sybase data server or Oracle instance.
AUTOSYSDB2 Indicates the name of the target Sybase data server or Oracle instance.
tee Indicates the command used to create a record of the autobcp output. It is
/tmp/autobcp.out recommended that you use the tee command to record the output of autobcp.
While autobcp is running, for each table in the target server, you will see autobcp
perform these steps:
1. Dump the table in the source data server to the output file you specified.
2. Delete the current data from the table in the target data server.
3. Load the data from the output file to the table in the target data server.
4. Dump the transaction log in the target data server (for Sybase only).
When autobcp has completed, you will see the following line:
"autobcp:Complete"
Bring up the event processor on the event processor machine using this
command:
eventor
If you are using a shadow event processor, use the following command to start
both the primary and shadow event processors:
eventor -M shadow_machine
The event processor marks both event servers as being in dual-server mode.
Client processes and commands check the flags in both event servers for
consistency. Therefore, it is imperative that you start the event processor before
you run any other commands.
It is safe to run the autobcp script while jobs are running on client machines. In
the worst case, some events on the first event server may not be written to the
second event server. This is not a problem because the event processor always
reads from both event servers. If it finds an event on one server that is not on the
other, the event server copies the missing event to the appropriate event server.
If one server misses an event due to recovery or spurious network problems, this
feature also dynamically synchronizes both servers.
■ The primary and shadow event processors and the third machine must all be
installed on the same type of machine, either UNIX or Windows.
■ Both event processor machines require installed event processors and remote
agents and valid server licenses.
■ The third machine requires an installed remote agent and a valid client
license.
■ The user who starts the event processor (the edit superuser) must have
accounts on all three machines.
■ Ensure that the configuration parameters are identical for both the primary
and shadow event processor machines (since the primary and shadow event
processors are typically installed on separate machines, and with separate
file systems for AUTOSYS and AUTOUSER). On UNIX, the
config.$AUTOSERV configuration file on both the primary and shadow
machines must contain identical settings for the third machine parameter.
■ Ensure that the primary and shadow event processor machines have all the
Unicenter AutoSys JM software, the AUTOSYS directories, as well as the
AUTOUSER directories, mounted locally. Otherwise, a problem with a file
server could stop both event processors.
Installation Steps
If these values are different, the /etc/auto.profile file must be edited on the
shadow machine, with the correct entries defined and exported. For
example, assume these are the values of AUTOSYS and AUTOUSER on the
shadow event processor machine:
/usr/vendor/autotree/autosys
/usr/vendor/autotree/autouser
Note: The primary and shadow event processors must be on the same
platform, either UNIX or Windows.
2. Install the remote agent on the third machine. To do this, perform a client
installation.
Note: The third machine must have a remote agent installed on it, and it
must have a valid client license. In addition, the third machine must be
installed on the same type of machine on which the primary and shadow
event processors are installed, either Windows or UNIX.
3. Edit the configuration file on both the primary and shadow machines to
specify the name of the third machine. For example, for a third machine host
named “kodiak,” enter this in the configuration file on both the primary and
shadow machines:
ThirdMachine=kodiak
4. To start the primary and the shadow event processors simultaneously, enter
the following command on the primary event processor machine:
eventor -M shadow_machine
Note: If you try to start the primary and shadow event processors without
having a third machine specified in the configuration file, the shadow event
processor will not start.
To restore the primary event processor, which is down, after the shadow event
processor has taken over, you must stop the shadow event processor, and then
restart the primary.
If you are installing multiple instances on UNIX machines only, you can use the
same remote agent port number for each instance. Therefore, you can use the
same remote agent to run jobs for all instances. However, if you are installing
multiple instances on both UNIX and Windows, you must specify unique remote
agent port numbers for each instance, as explained in the following section.
If you are running multiple instances across platforms, on both UNIX and
Windows, each instance must use different remote agent port numbers and
logging directories. You must do this because Unicenter AutoSys JM for
Windows requires that you install a remote agent for each instance. Therefore,
on Windows, each instance must have a unique remote agent port number, and
each must use the same port number on every remote agent machine. In
addition, each uses an instance-specific logging directory so that the remote
agent log files are separated by instance (AUTOSERV).
When you are installing on UNIX and want to run on both UNIX and Windows,
each instance must have the same configuration settings on every machine on
which it is installed. When you install a server, you must supply a unique
instance name and remote agent port number (“port number for auto_remote”)
for each instance, and you must supply the same values for all installations of
that instance. On UNIX, after you are done installing servers for each instance,
you can install a single remote agent on each client machine, then configure each
one to run jobs for multiple instances.
In addition, when installing event processors on UNIX, you must supply each
instance with unique AutoRemoteDir and AutoRemPort values, which are
located in the configuration file. When installing event processors on Windows,
you must supply each instance with a unique remote agent port number and
enterprise-wide logging directory values.
If you desire to do so, you can install a remote agent for each instance, supplying
each with unique remote agent port number during the installation process.
However, this approach will take up more space, and it is not necessary on
UNIX platforms.
Note: All instances that run jobs using one remote agent must have the same
type of event server, either Sybase or Oracle. A remote agent can only write to
one type of database.
If the event processor for an instance is on a UNIX machine, you should ensure
that the remote agent port number and the remote agent logging directory
settings in the configuration file are unique for that instance.
For example, if you were running a PRD instance with the port number of 5285,
you would modify the $AUTOUSER/config.PRD file in the following ways:
AutoRemoteDir=/PRD.tmp
...
# Port number of auto_remote
AutoRemPort=5285
...
You must ensure that the logging directory that you indicate exists on the remote
agent machines. For example, in this case, all remote agent machines that will
run jobs for the PRD instance must have a /PRD.tmp directory.
Note: If you are running with a shadow event processor, you must also modify
the configuration file on the shadow event processor machine.
For information on the configuration file, see the chapter “Configuring” in the
Unicenter AutoSys Job Management for UNIX User Guide. For information on
setting these configuration parameters on an Windows machine, see the chapter
“Administrator” in the Unicenter AutoSys Job Management for Windows User
Guide.
This section describes how to configure one remote agent executable on UNIX to
allow access from multiple instances on both UNIX and Windows NT. To
configure the remote agent, you modify the remote agent port number (socket
connection) and logging directory. Before you do this, you must determine
instance-specific port number and remote agent logging directory.
Note: Stopping the event processor and making these changes does not
affect the currently running jobs. These jobs will run to completion.
2. Modify the /etc/services file. For example, the file may have an entry like:
auto_remote 5280/tcp
To modify the file so that you can run both the ACE and PRD instances on
that remote agent, replace the existing entry with these two entries:
auto_remote_ACE 5280/tcp # AutoSys ACE
auto_remote_PRD 5285/tcp # AutoSys PRD
where:
auto_remote_ACE Indicates the service name for the remote agent, which you have modified to be
instance-specific. The remote agent service name must be the same in both the
/etc/services file and the /etc/inetd.conf internet demon configuration file.
5280/tcp Indicates the port number for the remote agent. If you wish to have multiple
instances running on one remote agent, you must change the number for each
instance designation. This number must be the same value across all remote
agents for that instance.
Note: If your machine is running NIS or NIS+, you must edit the services
file, remake the map, and push it out to all clients, including the server. For
information on how to do this, contact your system administrator.
To modify the file, include both the ACE and PRD instances by replacing the
existing entry with these entries:
auto_remote_ACE stream tcp nowait root /usr/vendor/autotree/auto_remote
auto_remote_ace
auto_remote_PRD stream tcp nowait root /usr/vendor/autotree/auto_remote
auto_remote_prd
Note: The specified directory must exist on all client (remote agent)
machines.
2. In the config.EXTERNAL file, add an entry similar to one of the following for
each instance for which cross-instance dependencies will be implemented.
■ For Sybase:
instance:EventServer=data server:database
■ For Oracle:
instance:EventServer=tnsname
where:
instance Indicates the three-letter instance ID for the target instance (for example, PRD).
data server Indicates the name of the Sybase data server used by the target instance. This
name is the same one specified in the configuration file of that target instance.
The data server name must be unique for each event server parameter entered
in the file, regardless of the instance with which it is associated.
database Specifies the name of the Sybase database used by the target instance.
tnsname Specifies the name of the Oracle data server used by the target instance. This
name is the same one specified by the event server parameter in the
configuration file of that target instance.
For example, if you were identifying a Sybase event server where PRD is the
target instance, AUTOSYSDB is the data server, and autosys is the database
name, the entry in the file would look like:
PRD:EventServer=AUTOSYSDB:autosys
Once you have created a config.EXTERNAL file for each involved instance, you
can define jobs to have cross-instance dependencies.
■ The config.EXTERNAL file can contain no more than 249 lines (or entries).
■ For instances that are running in dual-server mode, you must include an
entry for both servers in the config.EXTERNAL file. Furthermore, these
entries must appear together, consecutively, in the file; no other instance can
be inserted between them. For example, your configuration file may look
like:
ACE:EventServer=AUTOSYSDB0:autosys
PRD:EventServer=AUTOSYSDB1:autosys
PRD:EventServer=AUTOSYSDB2:autosys
XYZ:EventServer=AUTOSYSDB3:autosys
Introducing CAICCI
A
CAICCI (Computer Associates International Common Communication Interface)
is a transport layer that allows the asbIII process, which handles cross-platform
events, to communicate with Agents on AS/400, OpenVMS, UNIX, Windows,
and OS/390. On UNIX, CAICCI consists of several processes and a library. The
processes are responsible for network transmission, creating and maintaining
CAICCI resources. asbIII accesses the CAICCI API through the shared library.
■ caiccid
This process is referred to as the main CAICCI demon because it is started
first, builds the CAICCI resources and starts the other two CAICCI demons.
■ caicciclnd
This process is referred to as the clean demon because its responsibility
includes the maintenance of the CAICCI IPC resources.
■ caiccirmtd
This is the remote demon process, which is responsible for the transmission
of data across the network.
Installing CAICCI
The Unicenter AutoSys JM installer offers the option to install CAICCI with an
AutoSys server. If you select Unicenter CCI, it installs CAICCI automatically.
Later in the dialogs, the installer will ask for a list of remote hosts with which
CAICCI is to communicate. Enter their names separated by spaces. You may
change the list of remote hosts after installation by editing the file ccirmtd.prf as
described below.
Configuring CAICCI
The asbIII process communicates to Agent machines and Connect using CAICCI.
You need to configure CAICCI to communicate with a particular machine.
■ caiccid.prf
■ ccirmtd.prf
■ cciclnd.prf
Note: We do not recommend that you update the caiccid.prf configuration file
unless the file hits the max_recvrs limit.
caiccid.prf
The caiccid.prf file tells the main CAICCI demons what to do and specifies the
Max_Recvrs value.
where:
nodename Identifies the machine on which the Enterprise Management CAICCI demons
are running.
CAICCI Demons
The following parameters specify what the CAICCI demons do and specifies the
Max_Recvrs parameter:
On UNIX, the process that creates shared memory is the main CAICCI demon,
CAICCI. When CAICCI starts up it creates the shared memory segment;
therefore, CAICCI must know beforehand how large a memory segment to
create.
The value of nn determines the size of the shared memory segment because it is
the number of RVTs CAICCID creates. This has the effect of limiting the number
of application receivers. CAICCI is shipped with a default value nn=48. You may
need to change the default value to match your installation.
Each application requires at least one RVT and each unprocessed CCI message
requires another. On a busy server, you may need to increase the value of nn. An
indication that you need to increase the value of nn is if you receive the
CAICCI_E_FREERVTS error message. This message is displayed on the system
log. On a busy server, you may need to increase the value of nn to 200 or 300.
After the value of nn is increased, asbIII may not start up because CAICCI has
not started. This sometimes happens because CAICCI must protect access to this
shared memory with the use of a semaphore group. CAICCI needs to create a
semaphore group with a semaphore identifier for each RVT plus three extras. On
most UNIX platforms, the number of semaphore identifiers in a semaphore
group is governed by the SEMMSL kernel parameter. It is important to be aware
of the following rule when increasing the first number of the Max_Recvrs
parameter, nn:
SEMMSL >= nn + 3
If the Max_Recvr parameter is set to a value higher than allowed, it will default
to the maximum. Sometimes an application will hang or will be too busy to pick
up its messages. In a situation such as this, you will see the error message:
CAICCI_E_RECVBUSY Target [ ] queue is full, sender [ ]
The default behavior is for the sending application to sleep while waiting for
room on the buffer. This may work for an application using CAICCI but not for
the remote demon.
Customizing
2. You are then prompted for the names of the nodes with which the CAICCI
remote demon is to establish communications, as follows:
Please enter the name of the remote host or RETURN to end:
3. The prompt is repeated in order for you to specify another node with which
CAICCI will communicate. Respond with additional node names, one at a
time, until all of the nodes have been specified.
ccirmtd.prf
The ccirmtd.prf file identifies the local CAICCI node name, the UNIX host name,
and the block size for the local and remote machines.
where:
nodename Identifies the machine on which the CAICCI demons are running.
Syntax
LOCAL = nodename cciname max_msg_size [startup | nostart] [port=1721 retry=n]
REMOTE = nodename cciname max_msg_size [startup | nostart] [port=1721 retry=n]
where:
nodename Indicates the hostname that will be passed to gethostbyname. This can be any
name resolvable to the correct IP address and does not have to have logical
connection to cciname.
cciname Indicates the logical name CAICCI will use to identify this host. This name is
determined by the ca_uname function at install time and by ca_nodename
during runtime. These functions are the equivalent of uname –n. This name can
be as long as 64 characters but an alias must be used for names greater than
eight characters.
max_message_size Specifies the maximum buffer that CAICCI will send or receive over the socket.
It is a good idea not to adjust this. Each side of the connection may have this set
to different values, up to 32KB. The lesser of the two values is used.
startup | nostart Tells ccirmtd whether or not to initiate a connection. Sometimes you may only
want one side to initiate the connection. This is handy for a UNIX admin client
when many people power down their PCs at night. You can eliminate many
annoying messages when CAICCI is recycled during the night if the server
does not start connections.
-1—ccirmtd will start with a two second retry interval and double after each
unsuccessful retry attempt.
■ This is used in conjunction with the nostart option to allow the server to sit
passively and wait for incoming connection requests. If a client host goes
down the server will not attempt to reconnect, and we are again relieved of
messages requesting that we check to see if CAICCI is active on the client.
For example:
LOCAL = abcdef31 abcdef31 32768 startup
REMOTE = abcdef33 abcdef33 32768 startup
REMOTE = abcdef33 abcdef33 32768 startup port=7000
cciclnd.prf
The cciclnd.prf file defines the number of seconds to sleep between system scans
for communications buffer and connections cleaning. The default time value for
cciclnd is one second. The default value should not be changed unless instructed
by Computer Associates Technical Support.
where:
nodename Identifies the machine on which the CAICCI demons are running.
CAI_CCI_DEBUG
Definition:
Use:
Components Affected:
When to Use:
CAI_CCI_LOG
Definition:
Use:
Components Affected:
When to Use:
CAI_CCI_CONFIG
Definition:
Use:
Components Affected:
CAI_CCI_SHMMIN
Definition:
Use:
Components Affected:
When to Use:
CAI_CCI_PORT1
Definition:
Use:
CAI_CCI_PORT1=n
where:
n > 1024 K
Components Affected:
When to Use:
CCI_SELECT_TIME
Definition:
Use:
CCI_SELECT_TIME=n
where:
n >1
Components Affected:
When to Use:
Sometimes network conditions will cause the TCP/IP handshake to take a long
time to complete.
3. Remove the message queue for the caiccid process returned from Step 2,
enter:
ipcrm -q <message queue id>
4. Remove the shared memory for the caiccid process returned from Step 2,
enter:
ipcrm -m <shared memory id>
5. Remove the semaphore for the caiccid process returned from Step 2, enter:
ipcrm -s <semaphore id>
Platform Notes
B
This appendix describes platform-specific information.
C 4.5
If you are now using AutoSys and want to upgrade, see the Unicenter AutoSys
Job Management Upgrade for Windows User Guide for complete instructions
about how to upgrade AutoSys versions 3.x to AutoSys 4.5.
■ A backup of the database exists (if you are keeping one). If you have a
bundled Sybase database, the program will remove it. If you have an
unbundled Sybase database, or an Oracle database, the program will not
alter it.
Note: The uninstall program gives you the option of deleting your database.
To delete the database, check the Remove Database check box on the
uninstall window.
Removing
To remove Unicenter AutoSys JM, follow these steps on each machine where you
installed a Unicenter AutoSys JM component for a particular instance, select one
of the following depending on which database you used:
lsm -e UnicenterAutoSysJM-ORA
or
lsm -e UnicenterAutoSysJM-SYB
All of these procedures retain the $AUTOUSER directories for your Unicenter
AutoSys JM instances for future reference. To remove them do the following:
rm -r $AUTOUSER
You can mount the product CD and repeat the original installation command of:
./install-oracle.sh
or
./install-sybase.sh
The procedure listed previously, do not remove eTrust AC or the Unicenter CCI
since you may have licensed those products separately.
eTrust AC
$SEOSDIR/bin/secons -s
$SEOSDIR/bin/uninstall_eTrust
secons shuts down the eTrust AC daemons, and uninstall_eTrust removes eTrust
AC.
Unicenter CCI
export CAIGLBL0000
CAIGLBL0000=/opt/CA/CCISA
$CAIGLBL0000/cci/scripts/cshut
$CAIGLBL0000/scripts/setupCCISA -d
rm -r $CAIGLBL0000
cshut shuts down the CCI daemons, and setupCCISA removes most of the files
and rm removes the rest.
machines, 1-9
owner, creating, 3-2
profile, 5-10, 5-11, 7-10, 7-11
system architecture, 1-6
A
AutoSys Administrator, 1-12
AutoSys
CAI_CCI_Config, A-9
configuration file, 1-12
database, defined, 1-4 CAI_CCI_Debug, A-8
directory structure, 1-14 CAI_CCI_Log, A-8
environment, sourcing, 1-11
instances CAI_CCI_Port1, A-10
defined, 1-10 CAI_CCI_Shmmin, A-9
multiple, 1-18
Index–1
CCI checklist, 2-19
cciccid.prf, A-2 Client, 2-20
cciclnd.prf, A-7 Server, 2-19
ccirmtd.prf, A-5 cross-instance job dependencies, 1-20
Configuring, A-2 database verification, 4-18, 6-21
Environment Variables, A-8 file, 1-12
Introduction, A-1 in AutoSys Administrator, 1-12
RVT, A-3 installation:, 4-11, 6-14
Starting and Stopping, A-11
cross-instance configuration, 1-20
Troubleshooting Tools, A-8
CCI_Select_Time, A-10
ccicci.prf
D
Customizing, A-5
definition, A-2 database
defined, 1-4
cciclnd.prf, A-7
information, 1-12
checklist unrecoverable error, 1-25
client, 2-24
dataserver. See database, 1-4
existing Oracle database, 3-11
Oracle client, 7-2 DEC OSF/1. See Digital UNIX, 3-5
Oracle server example, 6-2
DG/UX
server, 2-21
TZ Variable Setting, B-1
Sybase client example, 5-2
Sybase server example, 4-1 Digital UNIX
unbundled Sybase database, 3-8 configuring for bundled Sybase, 3-5
client directory
checklist, 2-24 autosys, 3-2
configuring internet daemon for NIS and NIS+, autouser, 3-2
5-4 sadb, 3-2
installation, checklist example, 5-2 directory structure, 1-14
installation, Oracle, 7-1
installation, Sybase, 5-1 disk space requirement, 2-1
Oracle checklist, 7-2 DISPLAY variable, 1-11
client machine, 1-9 DSQUERY environment variable, 1-13, 3-9
client machines, 2-7 dual event servers
components, 2-4 defined, 1-4, 1-23
event processor, 1-5, 2-7 running, 1-25
event server, 1-4
remote agent, 1-5
E
config.$AUTOSERV file, 1-12
environment variables
CCI, A-8 I
DSQUERY, 1-13, 3-9
ORACLE_HOME, 1-13, 3-12
IBM AIX
SYBASE, 1-13, 3-9
configuring for bundled Sybase, 3-7
eTrust Motif 1.2 Requirement, B-1
user and host names, 2-8
identifying client machines, 2-7
eTrust AC:, 4-11, 6-14
identifying machines, 2-7
event processor PMDB, 2-10
defined, 1-5 policy manage host name, 2-10
machines, 2-7 policy manager user name, 2-9
See also event_demon
identifying machineseTrust Access Control, 2-8
eventor, and shadow event processor., 1-5
starting, 4-10, 6-13 identifying server machines, 2-7
Index–3
event server name, 2-15 Sybase, 4-1
gathering, 2-12 Sybase client
index tablespace (Oracle), 2-17 bundled Sybase preparations, 5-4
inetd file, 2-18 GUI display, 5-12
instance, 2-15 loading software on client machine, 5-5
monitor type, 2-14 low level, 5-9
owner, 2-14 Sybase server, 4-1
password (Oracle), 2-17 testing, 4-7, 6-10
password (Sybase), 2-16
Installation
primary database (Sybase), 2-15
configuring:, 4-11, 6-14
remote agent, 2-18
License:, 4-14, 6-17
remote agent path, 2-18
testing:, 4-16, 6-19
remote agent port, 2-18
remote command service, 2-13 instances
directories, 2-14 AutoSys, 1-10
installation directory, 2-14 cross-instance configuration, 1-20
port number, 2-14 multiple
valid hosts, 2-14 running, 1-19
sa user (Sybase), 2-15
interfaces file, 3-9, 5-10
system files, 2-14
tnsnames.ora (Oracle), 2-16 internet daemon
unbundled Sybase, 2-16 configuring for NIS and NIS+, 3-3, 4-13, 6-16
Unicenter CCI, 2-13 services, for NIS and NIS+, 5-9
installation directory, 2-13
remote hosts, 2-13
wizard setup, 2-12 J
installation
changes to files and directories, 1-17 java based listener, 2-6
client
Oracle, 7-1
Sybase, 5-1 L
directory, creating, 3-2
gathering information, 2-12 listener, 2-6
Oracle client compatability, 2-6
GUI display, 7-12
loading software on client machine, 7-4
low level, 7-9 M
overview, 2-1
rerunning, 4-7, 6-10
machines
script
client, 1-9
Oracle server, 6-8, 7-5
identifying, 2-7
Sybase server, 4-4, 5-6
server, 1-9
server
Oracle, 6-1 memory requirement, 2-1
N P
O
R
OPEN LOOK, 5-13, 7-13
Index–5
S requirement, 2-1
server installation, 4-1
SQL.INI file, 1-13
sadb directory, 3-2
unbundled, required database changes, 2-15
selecting components, 2-4 verifying database connection, 4-14, 6-17