writing the bulk data in JDE OneWorld

Narendra Gupta

Member
Hi,

i have to write the bulk data in the JDE OneWorld table. what are the supported API's and which API is best in performance.

if anyone have sample program to write, it will be very helpful for me to understand.

Thanks
Narendra
 
OK let's start with the easy questions.

What data?
How much data?
Which tables?
Do you need to go through the API for validation and logging?

It would be helpful if you set out your system configuration. A lot of people set it in their signature block so that its always available to others who want to help.
 
Hi,

Actually we are planning of migration from some existing ERP to JDE OneWorld application.so we have to write bulk data into JDE OneWorld application.we are evaluting the different approaches like table conversion, interface table, flat file conversion.

it would be great if anyone gives input on, which approach or API's are good in performance and their sample program to write into JDE OneWorld.

Thanks in advance.
Narendra
 
Hi,

Table Conversion and flat file conversion are pretty much the same : you export data from your own ERP to some .TXT
or .CSV file, those files are read by some custom JDE
Table Conversion program (written by you) and data is
written directly into JDE tables.

That approach can be quite risky because you directly
insert data without running any prior validation.

Interface tables (or also known as Z1) are temporary
tables which you have to fill up (typically by SQL
INSERT queries or massive database importing from
text files or by table conversions too).

Once those Z1 tables are filled up, you have to run
JDE batch processes that will validate their contents
(you'll run them several times until the validation is OK)
and finally will be imported into master JDE tables.

I personally prefer the Z1 approach, at least it provides
some validation on what you're inserting.

Most of this activity is disk-intensive (reading and
writing records), so I doubt you may have any performance
gain by writing the import process with C APIs instead of
going through easier JDE toolkit tools.
 
Hi,

Thanks a lot Sebastian.
If anyone have an idea about What are the API's by which we can write the data into interface table and the what are the API's to call the JDE batch process which can validate their
content and import into master JDE tables.it would be great if anyone can share sample program about how to insert into interface table and then how to call JDE batch process to import into master JDE tables.

Thank in advance.

Narendra
 
Hi Narendra :

Unfortunately, I don't have any document on those APIs,
it's far beyond my humble knowledge.
Regarding the documentation, you can download JDE dev
manuals from Oracle Web site.
 
[ QUOTE ]
Hi,

Thanks a lot Sebastian.
If anyone have an idea about What are the API's by which we can write the data into interface table and the what are the API's to call the JDE batch process which can validate their
content and import into master JDE tables.it would be great if anyone can share sample program about how to insert into interface table and then how to call JDE batch process to import into master JDE tables.

Thank in advance.

Narendra

[/ QUOTE ]

Forget the API's it will take way too long to get them to work properly for a one time conversion.

I'm assuming someone on the team has somekind of basic JDE Knowledge and your not writing blind to the JDE Master files.

the best way is to enter the information using the JDE front end screens, then check the databases to see what was written and what files/fields you need and what data you will be expected to populate. Your program should validate the information coming into the system before the records are writte to the file. I like my one time conversion programs to have two massive loops on the imported data. The first loop validates the entire set of data on the import. Item master, UDC Code, everything. all errors are reported. If one record is bad, then the second loop which writes the data to the master files never gets run. Meaning you must have clean data before ANYTHING gets written. Then run full simulations on the data. The more data the better.


Once program for each type of transaction or set of transaction.
One program for close orders
One program for open orders (if they cannot be key punched. )
One program for the G/L
One program for the A/P
One program for the A/R
etc etc..

One of the manuals has the flow charts to all the major system which gives you all the table names and how they relate to each other.

Rob
 
Back
Top