Invalid Structured Field Header in .ind file

[jpdevin]

 I see that others have reported this issue, under different circumstances.
I have a large amount of data to archive which was converted from another system. This data consists of ASCII text data that is being archived with Indexing parameters (EG: not generically). These are customer statements separated by day. For example, one day may have 15 statements, which are loaded by row breaking on a new account number. Another day may have a few hundred of the same.
 Am getting intermittent Invalid Structured Field Header in .ind file. There may be one or two out of a large group. Or with more frequency.
 It leaves the .ind files and .out files hanging which must be cleaned up. Have had failures, that then worked on second or third attempt without changing anything. Have other failures that keep getting the error.
 Initially, I thought it had to do with an Index field that I'm capturing, DATE; changing formats from a 2 digit year to a 4 digit year, but that doesn't seem like it's the case.
 The error message is somewhat non specific on how to diagnose and proceed -
arsload: Processing file >STMT_savaccountstatements_0001.20101227.txt<

arsload: 03/26/12 10:32:34 -- Indexing started, 371798 bytes to process
0425-415 CC=YES
0425-415 CCTYPE=Z
0425-415 CONVERT=NO
0425-415 CPGID=850
0425-415 MCF2REF=CPCS
0425-415 TRC=NO
0425-415 FILEFORMAT=STREAM,(NEWLINE=X'0A')
0425-415 TRIGGER1=*,72,X'504147453A2020202031',(TYPE=GROUP)
0425-415 FIELD1=1,55,15,(TRIGGER=1,BASE=0)
0425-415 FIELD2=1,72,10,(TRIGGER=1,BASE=0)
0425-415 FIELD3=14,23,35,(TRIGGER=1,BASE=0)
0425-415 INDEX1=X'6163634E554D',FIELD1,(TYPE=GROUP,BREAK=YES)
0425-415 INDEX2=X'7264617465',FIELD2,(TYPE=GROUP,BREAK=NO)
0425-415 INDEX3=X'4E614D65',FIELD3,(TYPE=GROUP,BREAK=NO)
0425-415 DCFPAGENAMES=NO
0425-415 UNIQUEBNGS=YES
0425-415 IMAGEOUT=ASIS
0425-415 INDEXOBJ=GROUP
0425-415 INDEXSTARTBY=1
0425-415 INSERTIMM=NO
0425-415 RESTYPE=NONE
0425-415 INPEXIT=/usr/lpp/ars/exits/acif/asciinp
0425-415 inputdd=/nnedwnld/011/Statements/STMT_savaccountstatements_0001.20101227.txt
0425-415 outputdd=./STMT_savaccountstatements_0001.20101227.txt.out
0425-415 indexdd=./STMT_savaccountstatements_0001.20101227.txt.ind
0425-415 resobjdd=/dev/null
0425-440 ACIF AT PK72203 HAS COMPLETED NORMALLY WITH RETURN CODE 0.
arsload: 03/26/12 10:32:34 Indexing completed
arsload: 03/26/12 10:32:34 -- Loading started, 371565 bytes to process
Invalid structured field header in the file ./STMT_savaccountstatements_0001.20101227.txt.ind
Loaded 0 rows into the database
arsload: 03/26/12 10:32:34 Loading failed
arsload: Processing failed for file >STMT_savaccountstatements_0001.20101227.txt<

[jeffs42885]

Check the file for hidden characters?

[jpdevin]

 Thanks very much for taking the time to reply/suggest. Unfortunately, some of these files are 100's of millions of bytes (importing customer statements from another system). And the error message listed doesn't give any sort of clue where the field being flagged as bad is. Plus the load may or may not work on rerun.
 So really at this point, I have no way to diagnose this vague error.

[Alessandro Perucchi]

Hello,

Well, with self created index the suggestion of Jeff makes sense.
Now I'm a bit puzzled, why the index that CMOD created is "wrong".

What is the version of CMOD that you are using? On which Platform?

Depending on the error, maybe the solution is in a fixpack.
Or not...
And maybe that could be a matter to open a PMR by IBM.

In all cases, you also have the option to trace what arsload is doing:

Code: [Select]

arsload ... -F T -L 15 -T /tmp/mytrace.log ...
Hope that helps a bit.

Sincerely yours,
Alessandro

[jeffs42885]

Bumping this back to life to see if there was ever a conclusion as to what the issue was?

I retrieved some files from production that need to be reloaded into another environment and I am seeing the same thing too- here is my command:

arsdoc get -u admin -h database-d /home/admin/ -G AG -c -g -i "WHERE DOC_NAME LIKE '%1FAA%'" -N -o AG.out

I got the index files and made sure that everything looked good, aka the INDEX files were pointing properly to the right file. I renamed the file so our daemon could read them, and now I am getting this error.

9.0.0.3, RHEL, Oracle, BTW

Could it be a codepage difference in environments? in prod, we are using 1208, preprod, 0

Thanks,

Jeff