DSXchange

I always worked with data stage reading from any database like Oracle, DB2, SQL Server, REdBrick, etc..

But now I will need read from Mainframe and as I don't have any doc to help me, I don't know where start by.

Can anybody help me?

Regards
Fernando

Fernando,

most of the time your host or mainframe data is going to get to your unix ETL environment in the form of flat files - either given to you by the operators or by allowing you to FTP the data from the host. The exception to this is DB/2 which can (if set up correctly) be accessed directly.

There is no such animal as a "mainframe stage" - I wish there were...

If you need to process in mainframe itself, you may need DS/390.

Sorry for delay, I was out of office and ...
Well, my question wasn't good.
Trying again.

I need to read a DB2/Mainframe and load a Oracle/Unix.
My Data Stage Server is in Unix.

And I don't know where start by.

For example:
1 - In this case when I create a job, I indicate that this new job will be a Job Sequence, a Server Job or a Mainframe job?


Regards
Fernando

Hi Fernando,

You can use a simple server job, But your actual job design depends on

1) How are you getting the data from mainframe?.

2) Is it through directly reading DB2 tables?

3) Are you being provided the mainframe data in a flat file or mainframe dataset?

You need to provide more details about your requirement, for us to be able to help you further.

Hello Fernando,
A good point to start will be looking into the source file which you will be getting. If it is a cobol file then identify the metadata with the help of .cfd. Once you understand the fields in the source field and define the metadata accordingly in datsatage then you will be good to go.
Thanks,
Ash.

Hi, Naveen.
Well, this is a good point.
At this moment I have to read from db2/mainframe.
Is the db2 stage able to read a db2/mainframe?

Regards,
Fernando

The easiest way to access DB2 data on a mainframe is via a 3rd party product such as DB2 Connect or IWay. These products provide easy connection from Unix and Windows users similar to ODBC connectivity. Talk to your local Ascential office about accessing DB2 mainframe data as they may have up to date information for you. It is not as easy as DB2 on AIX.

The next easiest way is to generate reports from your DB2 application and FTP them to you DataStage server for processing. DataStage can process delimited files or complex flat files generated by cobol programs.

The hardest way is trying to balance a DataStage MVS and DataStage Unix processing environment. You can get DataStage jobs running on MVS but you still face the problem of accessing that data on your Unix DataStage server so you can deliver it to Oracle. Probably easiest to take MVS out of the equation.

I've been at a number of sites who use Vincent's "hardest" approach happily. The mainframe jobs end in an FTP stage which transfers data files to the UNIX machine.

Ok, thank you!
Let me try now.

Regards,
Fernando

ray.wurlod wrote:I've been at a number of sites who use Vincent's "hardest" approach happily. The mainframe jobs end in an FTP stage which transfers data files to the UNIX machine.

Well, I think taht I will need to use Vincent's "hardest" approach happily.
Because the client here wants that I create Mainframe Jobs, and I don't have any doc to help me, I don't know where start by and I don't have Ascential suport because I'm a consult and second my project manager we can not use client's ascential support.

Can you inform me some sites that talk about this subject?

Regards,
Fernando Martins