I’ve downloaded and installed Natural for Windows Community Edition. I would like to run some Natural programs on it and read ADABAS files from our mainframe Development environment. How do I connect?
Also, is it possible to bring the FDIC file over to the workstation? I assume that I would need to download ADABAS for Windows Community Edition. I am somewhat familiar with using SYSOBJH to migrate objects / DDMs. How would I migrate the data from the mainframe to the workstation?
This is for a proof-of-concept. I’m pretty familiar with utilities on the mainframe side (ADAULD, ADALOD, etc.) but don’t know how to copy to another platform.
To access a mainframe Adabas database from Natural for Windows, you need to (purchase and) install Entire Net-Work.
Unless you are running Predict under Windows, there’s no need to bring over an FDIC. Your DDMs can be copied (use Object Handler with the TRANSFER option) to your FUSER (or FDDM).
To migrate the data from mainframe to Windows is a bit more complicated. If you have money to spend, there’s Adabas Event Replicator. The homegrown approach (simplified) is to decompress, FTP, and compress/load.
Thanks, Ralph. I need to use the SYSDIC-EL and SYSDIC-FI views so I can look at Predict data, so it looks like an install of ADABAS for Windows Community Edition, use SYSOBJH to transfer the two DDMS from the mainframe, then decompress / FTP / compress / load for the FDIC file as you suggest–very doable.
There have been discussion threads on the Developers Forum and on SAG-L that talked about the “gotcha’s” you should watch out for when moving code and data across platforms. They should be on your reading list; they will save you time and frustration.
Example; most people are wary enough to investigate anything having to do with sort sequencing. By contrast, the gotcha I have found ignored is the use of Hex constants (e.g. H’40’ mainframe is not appropriate on the PC, it must be changed to H’20’)
Good luck with the proof-of-concept. You will likely be pleasantly surprised at the run times possible on the PC; especially if you have a powerful PC and a very busy mainframe.
What I meant (didn’t say it very well), is that if you have a very busy mainframe, your response times and elapsed times probably suffer from the overall load.
By contrast, if you are running the same applications on a standalone, fairly powerful, PC, your times might not differ very much from the mainframe times.
