DSXchange

Hi All,

We are converting server jobs to parallel jobs. The trasformer
stage in the Canvas for Parallel Extender does not accept more than one input link (not even reference links). This is a big drawback since we cannot use the lookups(as reference links) from different hash files.

Regards

Hi All, I am updating/inserting in a table with a single server Job (There are two target stages for the same table) - one with an update on one condition and another with insert with another condition. Both constraints result in two different targets. But the table(not job) gets locked. Hence the j...

Set the jobs in the sequencer to "Reset" Mode. This would make the uncompiled jobs compile and run automatically

Regards

Ray,

This is a general query regardless of type of jobs(server or parallel).
Database used is Oracle 9i.

Sanjay: I do not find a Duplicate Stage in Datastage 7

Regards

Ray,VMC:

I am not using a hash file since it would only take a first matching record whereas a UV stage can take multiple records if multiple records match.

Hope you are clear about my problem

Regards

Hi All, I am joining 5 tables(using primary key columns of the 5 tables) to get the source data. What are the options you would use within this stage to ensure that this is correctly done i.e (without any duplicate data or cross product). I am presently using "distinct" to avoid duplicate ...

I am using an UV stage for inserting multiple records for a single record in the source . This turns out to be quite slow hence i have got the
multiple records in the source by doing a join (using sql). Is there anyway we can optimize the performance of a UV stage

Regards

Go to "CleanUp Resources" and try to click the "logoff" button after
selecting the appropriate job.
If this does not work then log onto the Unix box and kill the specified process


Regards

There could also be a problem with the OS settings. The C++ compiler
should be registered in the OS(AIX/Solaris/HP-Unix)

Please consult the "Install and Upgrade Guide" -- particularly page 6-14 - "C++ Compiler"

Regards

ODBC lookup should be at best avoided. Whenever possible we should
go for Hash lookup. ODBC lookup is much slower (probably more than 10 times slower) than hash lookup for records above 1,00,000.

Key columns in ODBC lookup can be matched using :1,:2... with
the columns on the source table

Hi All,

I want to create a file whose name would be dynamic. The file name has to be read from a sequential file.

Regards

Hi All,

I want to create a sequential file whose name would be dynamic. For example i want to create a sequential file which would have the current system timestamp appended to the file. Anyone having any ideas on this

You copy them in the *top* half, not the bottom. There is also a right-click option to 'Propogate Columns' to do something similar without opening the transform. You can select a column or columns and drag them over to the right-hand portion of the 'target' side of the transform - you will get an '...

chulett wrote:Errr... it is allowed. I'd suggest you try reinstalling your client as something seems to have gone awry with it.

copying columns which appear for source and target in the lower
half of the transformer (when u select properties)

[/b]When i try to copy input columns to output columns (by selecting the column in the input columns) in "Transformer stage" the datastage designer hangs.


This should be allowed since its required most of the time