I am trying to copy an adabas file to a flat file. It is huge and is taking very long. Would Read in Physical Sequence help?
Also, once the job abends for no space or because of too much time, how can I continue from the next record?
Thank you for you help!
Carol
Your reference to READ PHYSICAL implies you are using a Natural program for your extract. Likely it’s a logical read (READ BY or READ LOGICAL BY). READ PHYSICAL will be faster, but does not allow for restart. READ BY ISN performs better than READ LOGICAL but also allows for restart by including *ISN in the WRITE WORK statement. When the job fails, you can view the ISN of the last record, then re-run the job after changing the READ BY ISN to include FROM isn+1.
For READ PHYSICAL and READ BY ISN, the output should be considered to be in random sequence.
Regardless of which READ you choose, include MULTI-FETCH 10 for better speed.
Just curious. Why are you creating a flat file from the Adabas file? Is the flat file to be some sort of backup? If so, it would almost certainly be more appropriate to use Adabas utilities to save, and if necessary, restore an Adabas file.
Regardless of the answer to the above question, one addition to Ralph’s suggestions. Make sure that when you read from the Adabas file, you have a view with only the fields that will required on the flat file.