Most sites I have seen run CICS as a started task.
Most sites also control CICS and other on-line systems with an automated
operations package like CA-OPSMVS, IBM SA, etc. rather than a batch job
scheduler.
This is not universal just what I have observed as common practice.
This has the advantage of providing a clean virtual storage environment
to CICS each time it starts.
You also have the advantage that you do not have to define an initiator
to JES2 to run the started task especially at sites with hundreds or
thousands of CICS regions this would be a significant issue.
Mark Zelden provided the very nice answer below to a similar question at
search390.
The archives of CICS-L contain a number of discussions on running CICS
as a started task compared to running it as a batch job.
Thanks, Sam
What's the difference between a started task and a job? Can I run a user
program as started task? Mark Zelden
This question posed on 02 December 2002
Started tasks are generally used for long-running address spaces needed
for system functions (either "IBM operating system" functions, or other
software vendor product address spaces) or critical applications like
CICS, DB2 and IMS. A user program could be run as a started task, but
generally they are run as batch jobs. Started tasks may also have more
data set access and utility privileges (from a security standpoint), so
using them is normally controlled by protecting the started task
PROCLIBs defined to JES and requiring a special definition in your
security product (RACF, ACF2 or Top Secret) for the started task.
Here is some text from the chapter on started tasks from MVS JCL
Reference manual that will help to explain some of the differences:
When you determine where and when you want a set of JCL to run, you will
consider using batch jobs or started tasks. Batch jobs are scheduled by
a job entry subsystem (JES) and are scheduled to run based on the
resources they require and their availability, or based on controls you
put on the batch system. Controlling where and when a batch job runs is
more complex than using a started task.
A started task is a set of JCL that is run immediately as the result of
a START command. Started tasks are generally used for critical
applications. The advantages to using started tasks are:
You can control where and when your set of JCL is run. For example, you
can have the set of JCL started at each IPL of the system.
You can specify both static system symbols and JCL symbols in the JCL.
Static system symbols and JCL symbols provide additional control over
JCL that is used on different systems. For example:
- When access to production data sets is controlled to protect critical
business data, you can specify symbols that represent test data sets.
After testing the data sets, you can change the values of the symbols to
represent production data sets without changing the source JCL.
- When you need to swap in an older level of a subsystem while
diagnosing problems with a newer level, you can change the values of
symbols to represent the older subsystem without changing the source
JCL.
-----Original Message-----
A curious item (lore?) at one time it was "goodness" to start CICS with
a start command but in MVS there seems to me no real reason to do so
anymore.
Ed
<>
====================
This email/fax message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution of this
email/fax is prohibited. If you are not the intended recipient, please
destroy all paper and electronic copies of the original message.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
