DSXchange

Hi all,

I'm using a schema file to read/write into a sequential file. But when i delete the metadata in the stage, the job doesn't compile nor i'm able to view data.
My understanding is, schema files are used where metadata is not static and keeps on changing. Thus the metadata has to be changed only at one location and we dont have to load col defns in all the jobs/stages.
But if the job doesnt work without loading fresh metadata, what purpose does the schema file serve?
Please correct me if i'm wrong or tell me if i'm going wrong somewhere.

Thanks,
Chits

Hi Chits,
Its something like this schema file is used to import metadata into the stage during runtime.So for this you need to check runtime column propagation in the columns tab.To avail this option in your columns tab go to DataStage Administrator and check Runtime column propagation.Now when you check the runtime column propagation in columns tab and by specifying the schema file there is no need for you to load the column definitions.


Cheers,
Rajeev.

chitrangadsingh wrote:Hi all,

I'm using a schema file to read/write into a sequential file. But when i delete the metadata in the stage, the job doesn't compile nor i'm able to view data.
My understanding is, schema files are used where metadata is not static and keeps on changing. Thus the metadata has to be changed only at one location and we dont have to load col defns in all the jobs/stages.
But if the job doesnt work without loading fresh metadata, what purpose does the schema file serve?
Please correct me if i'm wrong or tell me if i'm going wrong somewhere.

Thanks,
Chits

Thanx Rajiv,
It worked.
But now i have another related query. In this scenario, if one has to do some transformations/calculations then how this can be done? coz now there wont be any columns visible in the transformer. I actually tried putting in a transformer in b/w two seq file stages. The job ran and simply dumped the data into destination file.
Or is it so that schema file can only be used for simple load/extract?

Thanks,
Chits

That's completely unrelated to the topic. Please start a new thread.

Hi Chits,
In the scenario you said transformations/calculations is not possible as the columns pass only during runtime from schema file and the osh code gets generated when you compile the job and that osh code is being ran when you run the job.If you want to do that then you have to load the columns explicitly even after specifying schema file and then use them in transformer for transformations thats the work around as far as i know.

Cheers,
Rajeev Katta.

chitrangadsingh wrote:Thanx Rajiv,
It worked.
But now i have another related query. In this scenario, if one has to do some transformations/calculations then how this can be done? coz now there wont be any columns visible in the transformer. I actually tried putting in a transformer in b/w two seq file stages. The job ran and simply dumped the data into destination file.
Or is it so that schema file can only be used for simple load/extract?

Thanks,
Chits