DSXchange

MetaStage has something called metabrokers which allow connectivity to various source, as well as some capability to push data back into those sources. No coding is required, but automation may be desired and that would require scripting. After you have all of your metadata gathered, you will write ...

Type 1, 2, 3, or a hybrid? How familiar are you with the types and their differences? None of them require usage of surrogate keys, but they absolutely make convincing evidence of solving problems. I suspect that you might not be totally familiar with their implementation, but the normal steps are: ...

Usually you have the originating key information, or at least a unique identifier stored with the data. In the case of surrogate key assignment, this is an absolute minimum. As far as auditing, you can perform reconciliations to show that all source rows are loaded via count comparisons based on nat...

Are you basically loading Oracle, but want to keep a synchronized copy in Teradata? What happens when the Oracle piece is retired in favor of just the Teradata side? You could consider making all transformation processes create insert and update files per table, that way you just need loading jobs f...

gbusson wrote:Which version does allow this feature?

7.5.x2

gbusson wrote:What are the requirements (hardware and software?)

Minimum is 1 cpu, 512MB RAM, Windoze XP Professional

May your billing rates be high, your runtimes low, and your temper even.

Cheers

The Stage icon allows multiple input and output links, but each link has its own filename associated to it. To have multiple links combine their owns in to a single stream, use the link collector stage. If you attempt to have multiple output links write to the same file simultaneously, you will end ...

Can you copy 1 row from your data file into a new file and then run against that? You should be able to massage that 1 row of test data until you isolate which column is causing the error. By the way, is this the first time you've tried loading into Informix using DataStage, meaning is this a new in...

Send your output links to a link collector and spool to the sequential file form the collector. Make sure the metadata is the same on all links. Each link will have its own constraint for its condition as well as applicable derivation for the reason output column.

You need to benchmark how long pure inserts take and how long pure updates take. Spool your inserts to a text file, spool your updates to another. Write a job that just reads the inserts text file and inserts into the target database. Copy that job and have it read the updates text file and change t...

Your metadata definitions represent the datatypes that the ODBC driver will bind to the returning rows. Inconsistencies here are best to be avoided. Importing metadata and using it insures that the database, ODBC driver, and DataStage all agree as to the content. User-defined queries are best to be ...

Or use the Aggregator stage and group by the key columns and use FIRST derivation for all other attribute columns. This effectively gives you the first row, as long as you handle NULL values properly, like with a substitute value to get it thru the Aggregator for non-nullable values (numerics) and t...

This is definitely weird, but just for grins, did you press the Get SQL Info button on the ODBC stage to make sure that the driver was verified for constructing the query? My assumption is that you could correctly import metadata and are using Auto-genered SQL. The only way you could get the trim() ...

Whooosh

A swing and a miss (to use a baseball analogy).

Don't know about AS/400 shared folders mattering but any chance m: is already in use? Meaning you can't detach because there's an open user/process in that file system?