|
View previous topic :: View next topic
|
| Author |
Message |
sprasannapathy
New User
Joined: 04 Jan 2008
Posts: 42
Location: india
|
|
|
|
Hi,
I am sincerely thankful the mainframe forum and all who helped me in understanding mainframe better.
Now, I have got a quick question for you. If I am using a VSAM file in //SORTOUT (SORT Utility). How to handle the duplicate keys? It is quite easy in IDCAMS by REPLACE parameter. Kindly help me.
Thanking you.
Prasannaa S |
|
| Back to top |
|
 |
Arun Raj
Moderator
Joined: 17 Oct 2006
Posts: 2481
Location: @my desk
|
|
|
|
sprasannapathy,
It would be better if you could explain your requirement further. You said your output file is a VSAM file, what are its attributes?; what about your input file attributes? If it's quite easy using IDCAMS, why dont you go for it? |
|
| Back to top |
|
|
sprasannapathy
New User
Joined: 04 Jan 2008
Posts: 42
Location: india
|
|
|
|
My input file is a PS file and it may contain key values that are found in the VSAM file (duplicates).
Yes, I am sure IDCAMS would do easy but the efficiency of SORT utility is much better than the IDCAMS. Now my question is there any option like REPLACE parameter to handle duplicate keys (as in IDCAMS) in SORT utility.
Thanks
Prasannaa S |
|
| Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006
Posts: 6250
Location: Mumbai, India
|
|
|
|
| sprasannapathy wrote: |
| is there any option like REPLACE parameter to handle duplicate keys (as in IDCAMS) in SORT utility. |
Word "handle" is quite generic in this context . . .what do you want to do with the records which comes up with "duplicate key"? Should they be removed, should they be written in seprate file, should they be just counted & represented in a form of report . . .or ..or.. or so on..
Suggest you start over and explain what exactly you are looking for, some one will be around with a better suggestion . . . |
|
| Back to top |
|
|
sprasannapathy
New User
Joined: 04 Jan 2008
Posts: 42
Location: india
|
|
|
|
Thanking you for reply.
While copying the input PS file to output VSAM by SORT Utility, if there are any duplicate keys in the input file then the duplicate key in the input file should replace the keys (data) in output VSAM file (that is what REPRO-REPLACE parameter do in IDCAMS). If you need anything further more please let me know.
Thanks
Prasannaa S |
|
| Back to top |
|
|
sprasannapathy
New User
Joined: 04 Jan 2008
Posts: 42
Location: india
|
|
|
|
Kindly help me........
Thanks
Prasannaa S |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix
|
|
|
|
Hello,
The reason there have been few replies is because what you want to do is not clear.
Suggest you post sample content of the vsam file before your process begins, a sample input ps file, and the content of the vsam file after the process executes. These do not need to be large samples - only enough to show your process clearly. The data might only be the key and some data field.
Keep in mind that your requirement is completely clear to you, but may not be to the others who read the topic. |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
| sprasannapathy wrote: |
| How to handle the duplicate keys? |
E38 User Exit? |
|
| Back to top |
|
|
sprasannapathy
New User
Joined: 04 Jan 2008
Posts: 42
Location: india
|
|
|
|
Sorry, for the late response.
Input PS file data
Key1 data1
Key3 data33333
Key5 data5
Output VSAM file
Key2 data2
Key3 data3
Key4 data4
When I am trying to copy the PS input file to output VSAM file using SORT utility job abended with the return code U0016. After the execution of the job when I did, look in to the output VSAM file I could see
KEY1 DATA1
KEY2 DATA2
KEY3 DATA3
KEY4 DATA4
The record Key1 data1 from the input PS file processed well.
I think while processing the record Key3 data33333, job got filed because of duplicate key. Now my question is how to overwrite the value Key3 data3 in the output VSAM file by Key3 data33333 (duplicate key in input PS file)
JCL for SORT
| Code: |
//*
//SORT1 EXEC SORTD
//SYSPRINT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SORTIN DD DISP=OLD,DSN=PSINPUT.FILE
//SORTOUT DD DISP=OLD,DSN=VSAM.OUTPUT.FILE
//SORTWK01 DD UNIT=SPOOL,SPACE=(CYL,(4,2),RLSE)
//SORTWK02 DD UNIT=SPOOL,SPACE=(CYL,(4,2),RLSE)
//SORTWK03 DD UNIT=SPOOL,SPACE=(CYL,(4,2),RLSE)
//SYSIN DD *
SORT FIELDS=COPY
/*
// |
SYSOUT Message
| Code: |
| WER255A VSAM LOGICAL ERROR 08 ON OUTPUT |
Attribute of input PS file
| Code: |
Organization . . . : PS
Record format . . . : FB
Record length . . . : 80
Block size . . . . : 27920
1st extent tracks . : 1
Secondary tracks . : 5 |
Attribute of output VSAM file.
| Code: |
Cluster: VSAM.OUTPUT.FILE
Data: VSAM.OUTPUT.FILE.DATA
Index: VSAM.OUTPUT.FILE.INDEX
Catalog: SYS1.CATLGU01
Type: KSDS
Data Index Data Index
Volume serial: TEMP93 TEMP93 Allocation % used: 1 3
Device type: 3390 3390 Number of extents: 1 1
CI size: 18432 512 Physical record size: 18432 512
CI % freespace: 0 Records per track: 3 49
CA % freespace: 0 CI per CA: 45 49
Allocation unit: CYL TRK Tracks per CA: 15 1
Primary allocation: 5 1 Number CI splits: 0 0
Secondary allocation: 5 1 Number CA splits: 0 0
Total allocation 5 1 Buffer space: 37376
Key length: 4 Total records: 4
Relative key position: 0 Records deleted: 0
Average record length: 80 Records updated: 0
Maximum record length: 80 Records inserted: 1
Records retrieved: 14 |
Thanking you for kind support.
Prasannaa S |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix
|
|
|
|
Hello,
A good practice is to resolve the duplicates before interacting with the vsam file.
Which sort product is used on your system? |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
Dick,
It's not dups in the input, it's the occasional already existing records on the output that he needs replaced rather than the more common add.
And why is this thread not in a sort forum? |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix
|
|
|
|
Hi CG,
| Quote: |
| it's the occasional already existing records on the output |
Yup, my reply probably could have been more explicit. I was thinking of unloading the current entries and getting a proper load file with no duplicates and then reloading the vsam file - assuming this had to be done with some sort/jcl.
Which is why the question about which sort product is being used. . .
If this was my requirement, i'd simply write code to read qsam file and update the vsam file as needed. . . |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
Yes, but since he can do it with IDCAMS, I wonder why he wants a sort solution.
Never the less, unless one of those who know better says otherwise, I think the only solution for him is an output exit.....
Who knows, maybe this might be a future enhancement...... |
|
| Back to top |
|
|
|
|
