Updates to Table Definition and its implications

[vijayrc]

Hi,
I have COBOL FD table definitions and LOAD in each of the STAGES. If I need to add another field to these table definitions, how would I go about reflecting the latest change to all STAGES that refer it?
Is the only option, going to all Stages that use these definitions, and manually do the update ? I think there should be a better way of doing this.. Help pls
Thanks, Vijay

[samsuf2002]

u can use RCP(run time column propagation)

[ray.wurlod]

The only good option is to use Usage Analysis to determine in which jobs the table definition is used. Re-import the table definition then manually edit the links where that table definition is used, loading the new column definition.

Following this approach means that linkages between your job's metadata and the table definition in the Repository are preserved, so that you will be able to use Usage Analysis subsequently (as well as Lineage and other analyses in MetaStage or its Hawk replacement).

[vijayrc]

The only good option is to use Usage Analysis to determine in which jobs the table definition is used. Re-import the table definition then manually edit the links where that table definition is used, ...

Thanks Ray. I would try out Usage Analysis to get a grip of changes that may be needed. Otherwise, is there any other mechanism to do this? I have a table definition used numerous time in a job, and can't imagine changing each occurance of it, when a change in Table Definition happens. Thanks in advance for your inputs.

[trobinson]

There are only the two mechanisms mentioned already. Either you have NO definitions in the job and stages by using RCP OR you develop your DataStage stages and jobs in a VERY structured, disciplined and methodical way with the use of Table Definitions. Using the latter means you maintain the table definitions and ONLY build/maintain columns from the Table definitions. It is extremely easy to cut the connection between the table definitions and their usage in stages.

[kumar_s]

If it is CFF stage and the file is Ebcidic, no other go, you need to under go this pain. But if its Sequential file, and if the added field comes to the last, and if the last column of the exisiting fiels are varchar fixed width, and if the additional colum is not required by other jobs or tables, you can skip it changes in other jobs.
But I dont guess you scenario will pass through all these 'and if's.

[ray.wurlod]

The only good option is to use Usage Analysis to determine in which jobs the table definition is used. Re-import the table definition then manually edit the links where that table definition is used, ...

Thanks Ray. I would try out Usage Analysis to get a grip of changes that may be needed. Otherwise, is there any other mechanism to do this? I have a table definition used numerous time in a job, and can't imagine changing each occurance of it, when a change in Table Definition happens. Thanks in advance for your inputs.

You'd better start imagining it, because it has to happen.

[Chuah]

Hi,
I have COBOL FD table definitions and LOAD in each of the STAGES. If I need to add another field to these table definitions, how would I go about reflecting the latest change to all STAGES that refer it?
Is the only option, going to all Stages that use these definitions, and manually do the update ? I think there should be a better way of doing this.. Help pls
Thanks, Vijay

Hi,

Use RCP with the "readcobol" utility provided by PX.This will enable you to dynamicaly source the COBOL copybook at runtime.
From your job, you call a unix script that rund readcobol to convert your COBOL copybook into a Orchestrate Schema file. In your job define a sequential file stage that has a Schema file defined.

Rgds,
Chin