DSXchange

Can you use a server job (or even a server Shared Container in the parallel job)? Parallel job Sequential File stage has ... problems ... with embedded quotes. Server job Sequential File stage is far better at handling them.

Can you use a protocol like Samba to see the disk of one machine on the other?

Otherwise, provided the permissions are all OK, there's no reason not to use an FTP stage to read the remote file contents into your job.

Thanks for helping future noobs to avoid this one!

Parameter values are read when the job starts. There is no legal way to change the value once the job has started.

Why so many "what is the difference between..." posts?

Enterprise stages have innate parallel capability (even if it can't always be gainfully employed). API stages originate from the server engine and always execute in a single (sequential) process.

That value is a key to the DS_LICENSE table, so I'm not sure where the FILE 29 reported came from. The key contains the MAC address and identity of the client machine and the process ID of the corresponding agent process on the server.

Yes it needs a column in the DataStage job. Otherwise there's nowhere that the DataStage job can put it after it's retrieved. The column data type should be the same as the column's whose MAX value is being reported.

DataStage product roles (there are four of them) are assigned to DataStage users via the DataStage and QualityStage Administrator client on a per-project basis, on the Project Properties dialog Permissions tab.

Parameters are stored in - as properties of - individual jobs. A Parameter Set is a re-usable component stored in the Repository that can be loaded into as many jobs as required. A Parameter Set contains a collection of ordinary parameter definitions. A Parameter Set may also have associated with it...

Thanks for the suggession... I am not sure how many output link i would require and also the number of recors within the same row is not fixed??? Is there any other way to achieve it?? Thanks for you help!! Not more than 10 "A" records is possible and not more than 6 "B" records...

My guess is that there is a problem for the login/security service writing to one of the metadata repository tables. You probably need to involve your official support provider.

First off, float and dfloat are poor choices for keys because they can not be guaranteed to be stored accurately. Char or VarChar are probably better choices here. BigInt if you must. The numeric results of month and day and hour and min may be single digit. You need to assure that they get leading ...

No, it's not additional characters, it's multiple bytes per character. In Oracle you can declare string lengths as BYTE or CHAR. Looks like yours are BYTE.

[In parallel jobs only...]
There is a button on the Format properties called Defaults> (or something like that). This allows you to set the current format settings as default, replace the current with your most recently saved default, or to restore factory settings.

Doesn't Oracle limit VarChar data types to 2048 characters, forcing you to specify Long VarChar beyond that?