DSXchange

Why do you have 16 output files?
The Teradata Enterprise stage invokes FastExport which produces 1 answer set. All the parallel processing is performed on the Teradata Server not in DataStage.
For 1.4 M rows I've seen better throughput using the Teradata API stage.

Bill:
You need to consult with your Teradata DBA concerning where tables are created, when they're created, access rights, etc...
Typically you should be qualifying the table name with a database name otherwise the DBMS uses the default database of the userid.

Here is the ecase assigned for the fix to this problem: ecase #58139.

This ecase was created 8/2004. I'm surprised it wasn't fixed in v7.5

This is a deficiency in DataStage. The compiler does not like the $ and produces the error.
You can not use an environment variable in the transformer stage it has to be a job level variable.
I believe I reported this problem to Ascential but I can't find the case # or their response.

We ran into a problem with the Oracle sparse lookup when returning a numeric field for matched records the value was always zero. I just received a patch from Ascential support to fix this problem but have not applied it yet. The patch is for v7.1. If the field being returned is char then the proper...

The 16 requested sesions is a rule of thumb I use for a Teradata system with more than about 100 VAMPS. Because the default setting in Fastload, MultiLoad, and FastExport is 1 session per VAMP I try to limit the number of sessions that my job starts to reduce start-up time. The sessions per player I...

It would be unusual to have a 7 AMP Teradata system. The number of AMPS is usually an even number; 7 nodes would be more likely. This would translate into x-number of AMPS; actually VAMPS. When the Teradata Enterprise stage is used for writing it runs the Teradata FastLoad utility. The default numbe...

This is a known problem in DataStage v7.1: 9 WARNING Tue May 24 09:40:47 2005 Teradata_Enterprise_12,2: TeraSession: cli rejected buffer write, stall #0 of 0. [terasession.C:1001] 10 FATAL Tue May 24 09:40:47 2005 Teradata_Enterprise_12,2: TeraUtils:Put failed: Error Code = 204 [terautils.C:2631] Ec...

You need to look at the Orchestrate doc to get info on this. Here's a snip from the manual. You would code the sessionsperplayer and the requestedsessions parameters on the connection string in the Teradata Enterprise stage but you have to select user defined connection string. This is from the Orch...

There is a version of DataStage EE scheduled for GA release on 3/31/05 that supports Suse Linux Enterprise(SLES) v9.

You need to discuss multiload functionality with your DBA.
You can not run multiple multiloads against the same table concurrently.

I just went through this problem with Ascential support. It requires an upgrade for the PXEngine.
I believe that the problem has to do with columns that were used as sort columns and not carried through the remaining stages.
Contact Ascential support for the fix for ecase 50039.

You do need the Teradata Client utilities installed on the DataStage server. Of course you need them for doing any Teradata access from DataStage; or at least the Teradata ODBC driver with a DSN configured. Here is the procedure: In Designer, right click on Table Definitions, click Import, click Plu...

I've successfully used the PlugIn Import feature from Designer to import a Teradata table definition.

Based on how PX creates UNIX processes, one for each parallel stage X the number of nodes, should there be a limit on the number of stages in a PX job and what should it be based on?