IBM Mainframe Forum

hi,
please tell me how to remove the duplicate records from the input file and write it into another dataset.I need to write the duplicate records seperately how to achieve this please tell me.



thanks

why is 0003333 there? It only appears once. Have you tried looking in the manual for DISCARD or SAVE or 'saving duplicates' or something like that?

I have used xsum and its working fine thanks...

Hello,

I believe xsum is a Syncsort function - not a DFSORT function.

Look at the output from this sort that works with XSUM and post the messages (with the message ids) here and we can see which sort product you are using.

If your system uses Syncsort, you need to post in the Syncsort part of the forum, not the DFSORT part of the forum.

Glad to know that. I was thinking of chucking XSUM into the mix but I thought it may have been a SYNCSORT keyword and I had no time to check in the manual.

Hi Nic,

After posting that, i peeked into the Syncsort doc - yup, XSUM is for Syncsort.

Then i went and looked for where XSUM is mentioned in topics on the other forum. All of the XSUMs in a sort control statement were Syncsort topics. Unfortunately lots of folks use a ddname of XSUM or SORTXSUM in DFSORT processes. . .

Maybe TS will provide the info.

Hi Dick, as I said I did not have time to look up manuals as I was on my lunch break. And I have never used SYNCSORT so I have never looked at their manual. And what is available is, anyway, dependent on the sort level and PTFs applied.