I would like to take the file created by ADAULD, uncompress it with
ADACMP, perform manipulations on it, recompress it, and load it using
ADALOD.
Can anyone provide a description/documentation/pointer to
documentation on the structure of the file?
Format of the decompressed file is described in the documentation, under Utilities, ADACMP - for example ADA74:
http://techcommunity.softwareag.com/ecosystem/documentation/adabas/ada744mfr/adamf/util/adacmp-DecOut.htm#adacmp-DecOut
You will need the FDT description in order to interpret the decompressed output.
There is an excellent third-party tool that is perfect for performing the task you are describing. The product is called ADAREORG and it is made by CCA in Australia. It uses the ADABAS ADACMP module for the decompression and recompression of the data, so you do not need to be concerned about supportability by SAG if you decide to go this route.
You can perform various manipulations in the middle step after decompression. I have used this tool to remove fields and groups from the file, rearrange fields, add or remove fields from PE groups. and to create new fields with pre-populated values. This product simplified the whole process and is worth its cost as a labor-reducing tool, let alone performance improvements from changes to data storage.
The CCA products have various distribution channels worldwide. Please view the product information at http://www.ccasoftware.com.au/products_adareorg.shtml and if interested, contact your local distributor as found at http://www.ccasoftware.com.au/home_world.shtml.
I have looked at the documention you referred me to at http://techcommunity.softwareag.com/ecosystem/documentation/adabas/ada744mfr/adamf/util/adacmp-DecOut.htm#adacmp-DecOut .
It says there that each record is preceded by a 2 byte record length and 2 bytes of zeroes.
I did ADACMP DECOMPRESS without the ISN number, and I am working on a VSE machine. I then FTP it to a unix machine. When I do a dump of the data, it starts from the first field, there is no recordlength or zero bytes preceding the data.
Any explanations? Is this a glitch in the documentation? Does it act slighly differently on VSE? Or have I missed something else?
I have looked at the documention you referred me to at http://techcommunity.softwareag.com/ecosystem/documentation/adabas/ada744mfr/adamf/util/adacmp-DecOut.htm#adacmp-DecOut .
It says there that each record is preceded by a 2 byte record length and 2 bytes of zeroes.
I did ADACMP DECOMPRESS without the ISN number, and I am working on a VSE machine. I then FTP it to a unix machine. When I do a dump of the data, it starts from the first field, there is no recordlength or zero bytes preceding the data.
Any explanations? Is this a glitch in the documentation? Does it act slighly differently on VSE? Or have I missed something else?
There are technical tips in ServLine24 that are relevant to your quiestion. I think Document ID 48001 is closest:
http://servline24.softwareag.com/SecuredServices/KCFullTextASP/viewing/view.asp?KEY=0048001-1004884&DSN=SCORE&DST=TCD&HL=1&QUERY=migrate|data|mainframe|unix&SessionID=188336453
Note that you will need to FTP the file in binary to the Unix machine. See the ADACMP options for RECORD_STRUCTURE and SOURCE_ARCHITECTURE. For earlier versions of Adabas, the tool cvt_format was supplied in the Tools directory. (See ServLine24 Technical Tip - Document ID 44293 at http://servline24.softwareag.com/SecuredServices/KCFullTextASP/viewing/view.asp?KEY=0044293-14688955&DSN=SCORE&DST=TCD&HL=1&QUERY=migrate|data|mainframe|unix&SessionID=188336453 if you need this)
Hi,
I can´t acess the link http://servline24.softwareag.com/SecuredServices/KCFullTextASP/viewing/view.asp?KEY=0044293-14688955&DSN=SCORE&DST=TCD&HL=1&QUERY=migrate|data|mainframe|unix&SessionID=188336453 .
The error is: you don´t have permission to acess
I need to transfer an adabas file from mainframe to Risc, AIX So.
I make a adauld and adacmp decompress on mainframe and a compress on aix, using the options record_structure=rdw
source=(ebcdic,high)
But when the compress executes, i get the error for packet columns.
What the best way to transfer files with packets format from mainframe to unix ?
Sorry for my english 
You need to specify the BINARY and RDW options in your FTP step.
I using the options SITE RDW and BYNARY on ftp …
Hello Sandra,
Compression Errors with Packed fields usually indicate that the record structure is shifted.
There are many possibilities why this might happen:
The ADACMP produces an error output that can be interpreted with the ADAERR utility.
ADAERR d dump produces an hex print of the input record and the offset where ADACMP detected the error.
You might have to count through the record to determine the cause of the error.
Of help could be an interpreted print of the record on the mainframe (ADAICK DSCHECK FILE=n,ISN=x)
Kind regards,
Marbod
Thanks very much !
The problem is that 1 generate the adacmp cards on mainframe, but on the physical file the sequence is another
Good that you found the cause of the error!
If the sequential input for the ADACMP is not in the FDT order you could use the following:
On Open Systems the ADACMP has a FIELDS parameter which allows to specify a different order or subset of the fields of the FDT.
On Mainframe the equivalent parameter is FORMAT which can also be used in the DECOMPRESS function.
Kind regards,
Marbod
I’ve written a utility, years ago, simplifying the transfer of a mainframe file to OpenSystems,
it can be found in (and downloaded from) this thread.
Note:
this has been written in the era of Adabas 7.x, so it does not know anything about the features introduced with Adabas 8,
if anyone needs specific Adabas 8 features supported … please let me know and I’ll see what I can do.