First let me say I have successfully loaded several of our files from the IBM to the Linux world. I have also had a few stubborn ones. There are two separate files I am struggling with currently and could use some help with. With one file I get an ADACMP error, tells me the adabas internal id which is ‘BR’ and that it has ‘invalid content for specified format’. I tried copying that file to another file on the mainframe and download the copy file, and amazing it works. Exactly the same file layout. Any Ideas?? The second file gives me an error in cvt_fmt says: mismatch between DATA and FDT detected, reason: MU count too large (240), This file has an Mu that only has 30 occurrances… any ideas on either scenario would be appreciated.
John
MU count too large (240), This file has an Mu that only has 30 occurrances.. any ideas on either scenario would be appreciated. Maybe that has to do with the little-endian and big-endian-logic. Aren’t you big endian on mainframe? I that case you have to tell ADADCU the TARGET_ARCHITECTURE. See ADABAS-Documentation for more information.
Well, sorry it took so long to post back, but we had other projects to tend to and I wasn’t able to continue with this one. However I am back on it now, and I have solved this particular problem. The culprit was a shortage of memory in the IBM virtual machine which left the FTP process short of space. I ran the FTP / CVT_FMT in a different environment and it worked just fine. SO, If anyone has an IBM running VSE-ESA, this may affect your plans to migrate to Linux as well. Check the memory with a GETVIS command. In our case we had to Re IPL the virtual machine.
John
To be more specific, the FTP actually had an error due to the memory shortage, which truncated the file, which in turn causes the CVT_FMT to have an error due to the aborted/truncated file. Hope that helps.
John