Adabas/Natural to Linux/Unix/Windows


We are looking for migrating/re-platforming from Natural/Adabas from Mainframe to Linux/Unix/Windows. The applications has around ~200 JCLS ,~ 3000 Natural programs and ~50 Adabas Tables . I have gone through some threads and comes across challenges related to Collating sequences , endian settings etc.

Few questions :-

  1. Are there any other issues that will also need to be taken care of . I see references to some old posts but those have been removed ?
  2. As I read endian settings depend on the processor so is there really a need to take care of ending settings when moving from one operating system to another?
  3. For Data migration from Mainframes to LUW what are the things that needs to be considered?
  4. Is there a way to calculate the effort that will be required to port this application from Mainframe to LUW?

Thanks in advance.

Virendra .

If a Natural module REDEFINEs an I2 or I4 field, results will be unpredictable; there will be no run-time failure. I have not seen a lot of this, but you still need to check your application and correct where necessary.

Thanks Ralph . By any chance can you provide insights for questions # 3 and # 4.

Thanks in advance.


Hi Virendra,

Some points to be considered:

  • ASCII/EBCDIC: Digits are < letters on ASCII, on EBCDIC vice versa.
  • Natural B fields should be defined as Adabas B fields with option HF on LUW.
  • Problems with field redefinitions have already been mentioned.

You are not the first one doing a migration from mainframe to LUW. I would recommend to ask our support if they can give you some tips.