Large .txt file with No indexing

[B.Rivers]

I have a large data file that users want stored in CMOD.  We defined the file using Generic indexing, but there is nothing to create an index on, since it is only a data file.  We stored using an .ind file.  The size is 314,572,800 kb.  It takes too long to open to see the file in the viewer.  About 10 Minutes.
I would like to use ACIF, with an index and make it a large object, but there is nothing to index on?

No sure how to proceed,   

Or is this just not what CMOD is used for.

Brian

[Greg Ira]

Did you try indexing only using constants?  We've done that a number of times and it doesn't involve pulling indexes from the document.
Load date is a good one to start with as you only need to specify the t in the load information.
Usually we throw in a title or report name for good measure.  Looks like this:
CC=YES
CCTYPE=A
CONVERT=NO
CPGID=500
MCF2REF=CPCS
TRC=NO
FILEFORMAT=RECORD,261
FIELD1=X'C8C5F8F9F9F6'           /* HE8996  */
INDEX1=X'D985979699A36DD5819485',FIELD1,(TYPE=GROUP,BREAK=NO)   /* Report_Name */
DCFPAGENAMES=NO
UNIQUEBNGS=YES
IMAGEOUT=ASIS
INDEXOBJ=ALL
INDEXSTARTBY=1
INSERTIMM=NO
RESTYPE=NONE

[Justin Derrick]

Enabling large object support for this report will help with the time between clicking and loading -- also, increasing object size and number of pages per large object will help with performance as well.  Finally, ensure that you're using OD77 compression to maximize storage space and minimize the volume of traffic going across the network.

-JD.

[B.Rivers]

Thanks for your sample.  I did get this to work!  Mine is a bit different because it was an ascii file.

(TYPE=GROUP,BREAK=NO) was key from your example  Thanks!

CC=NO
CCTYPE=Z
CONVERT=NO
CPGID=850
MCF2REF=CF
TRC=NO
FILEFORMAT=STREAM,(NEWLINE=X'0D0A')
TRIGGER1=*,21,X'2020',(TYPE=GROUP)                             /*            */
FIELD1=2,4,9,(TRIGGER=1,BASE=0)
INDEX1=X'E28583A3899695C9C4F1',FIELD1,(TYPE=GROUP,BREAK=NO)   /* SectionID1 */
DCFPAGENAMES=NO
UNIQUEBNGS=YES
IMAGEOUT=ASIS
INDEXOBJ=ALL
INDEXSTARTBY=0
INSERTIMM=NO
RESTYPE=NONE

[Justin Derrick]

Just out of curiosity, how many pages long is this report?  The largest single report I've seen was 80,000 pages -- and they used to print that monster out every week! 

[jsquizz]

Just out of curiosity, how many pages long is this report?  The largest single report I've seen was 80,000 pages -- and they used to print that monster out every week! 

One of my past projects we would load a massive text file weekly, I think it was 200k+ pages of numbers, indexing with a date trigger and a floating trigger on one of the line items. It would take DAYS to index. Half the time it wouldn't even load. I can't imagine retrieval. I went to my friend on the business side of the house and he even told me .."Nobody is going to retrieve this"

I've also seen 50k+ metacode files that would be converted from Xerox, to AFP, then to PDF, at load time. This was not fun, especially when I had to migrate from Xenos to AFP2PDF. Ironically, Xenos loaded it way faster! It would take maybe 2 hours to index with Xenos, but with the whole new "Conversion process/Integration" it would 5-6 hours if not longer. Fun times.