My intention is to compare two files by matching a key field from both files and create an output file containing matching records only.
1. The two files are VB FILES
ICE084I 0 EXCP ACCESS METHOD USED FOR T1
ICE084I 0 EXCP ACCESS METHOD USED FOR INPUT1
ICE218A 6 77 BYTE VARIABLE RECORD IS SHORTER THAN 94 BYTE MINIMUM FOR FIELDS
ICE751I 1 EF-K10929 F0-K30362 E8-K44563
ICE052I 0 END OF DFSORT
The base file - IHC$017.BCNJ.NASCO.COM.SUBMSTR is a vb file file but all the records in the file are of length 77(including 4 bytes length field).
Can you please help me why this happens even after I use ' VLENMAX'.
Joined: 15 Feb 2005 Posts: 7129 Location: San Jose, CA
This has nothing to do with the SPLICE operator or VLENMAX.
If you look at the TOOLMSG messages, you will see that the error occurred for one of the COPY operators, not for the SPLICE operator.
The problem is that your OUTREC statement has 35:35:60 which requires that the VB input record be at least 94 bytes, but you say the records in your input file are only 77 bytes. I don't know why you're using 35:35,60 when the records only have 77 bytes, but that results in a short record situation and causes the ICE218A.
Since I don't know exactly what you're trying to do or why you want to use 35:35,60, I can't tell you for sure how to do what you want, but you can fix the ICE218A problem by using a DFSORT/ICETOOL job like the following:
Hi Frank...Thank you so much for this...This time it worked.
Actually the the input and output files are similar variable length files. In each file there will be different types of data. Some types will be 77 bytes length..others had 120 bytes , 550 bytes and even 1020 bytes. Different segment records in the VB files thus has different lengths...
Also I would like to know how the issue was resolved...I find two chages here ...1. using common work file T1 2. Using BUILD . Which among these two resolved the issue and how...
Joined: 15 Feb 2005 Posts: 7129 Location: San Jose, CA
The IFTHENs resolved the issue. When you use IFTHEN/BUILD, it pads short fields with blanks. So that eliminated the "short" records.
Using concatenated temporary data sets with referback like you did can result in problems because of a system restriction. It's better to use one T1 MOD data set and avoid that kind of concatenation.