IBM Mainframe Forum

Hi All

What is the Max number of files we can override in a JCL.
I have an i/p file with 10 concatenated files in Job.


My job runs 12 times every day.When any job abends because of some issue,I need to include 10 input files of abended job in next job by giving overrides to INPUT
The next job will have 20 input files concatenated.
what is the max number of files we can override like this.Is there any limit?
When I searched I found as
'3273 is the limit of dd statments which can be coded in a single JCL so its a limit of datasets which can be concatenated
Please correct me If am Wrong'.
So I am not sure if that is the correct count.

Can you please let me know if the above is correct or do we have any other limit?

Thank You

Each DD statement in a concatenation counts as 1 DD statement for the step limit. The limit you quoted may be optimistic for two reasons

• Many shops use a smaller TIOT (which is what defines the limit); something like 1637 DD statements would be the limit for the default 32K TIOT
• Some DD statements require more than the normal 20 bytes for the most common allocation.

I don't know of any limit (beyond the limit for total DD statements) for override in JCL.

These limits are for a single step. Jobs with multiple steps can have more DD statements spread between the steps.

I suggest your terminology might be an issue. You mention 'overrides; but code it like in-stream JCL So, are you just asking how many DD statements you can code. The answer given is loads more than you will ever need.
You might want to consider coding it a different way.
You could make the input file a GDG with your //INPUT DD DSN=gdg base name, then at the end of the job (if successful, either delete all generations, or copy to a backup and then delete). By coding just the GDG base name you will automatically concatenate all generations, but be aware it starts with the LATEST generation, and works backwards.

halfteck wrote:...
You could make the input file a GDG with your //INPUT DD DSN=gdg base name ...

Won't work: each generation becomes what amounts to a DD statement.

well if he is allowed 000's of DD statements in a step, i am at a loss to understand why would it not work

There is a maximum of 255 datasets cataloged in a GDG at one time. See the manual, not the thousands you seem to imply..