DSXchange

It sounds like a bug. You should report the problem to IBM support. The connector calls setUnscaledByte in cases where you are selecting a DECIMAL column that has a length of only 1 byte, i.e. a DECIMAL(1) or DECIMAL(2). As a work around, you could try casting the column to an INTEGER or a larger de...

Have you tried using the Teradata API stage using the User-defined SQL option? Put your EXECUTE statement in the User-defined SQL.

It appears to be a combination of a Teradata Connector bug and a limitation in the Parallel Transporter's Update operator. If you update a table, missing rows and unique constraint violations go into error table 2 (the _UV table), and that error table will contain the column values that caused the v...

The Teradata Parallel Transporter API has a hard limit of 64260 for its buffer size. You cannot increase it (although you can decrease it).

Unfortunately, no. While MultiLoad does have a TABLEWAIT option that allows you to make one MultiLoad wait for the other to finish, the TPT API used by the Teradata Connector does not support such an option. If you want to use the MultiLoad utility, you'll have to use the Teradata MultiLoad stage an...

There is a property in the Teradata Connector called Array size. Reduce its value to get rid of the warning.

There is another way to get rid of that error that says, "Field ARGFIN is unreferenced in the SQL statements". As you discovered, the problem was that your lookup key contained field ARGFIN, but your SQL did not contain ORCHESTRATE.ARGFIN. In the connector properties, there is a property c...

You cannot have two MultiLoad utilities or TPT Update operators loading the same table simultaneously, because MultiLoad and the TPT Update operator lock the target table (which is why you got error 2574 : Table being MLOADED). You must either use the TPump utility, the connector's Stream operator, ...

If possible, try using the Teradata Connector. You can load the table using the Immediate access method, and it will be faster than TPump or the Stream operator, since you can use arrays.

Have you considered running the SQL via the Teradata API stage instead of using a DS routine?

The other option if you prefer to use the Load operator (FastLoad) is to load a temporary staging table, and then specify an INSERT...SELECT statement in the After SQL property to copy the data from the staging table to the target view. You can create the staging table using the Table action property.

You are not using the Update operator (MultiLoad), you are using the Load operator (FastLoad). Check the Load type property in the Teradata Connector and set it to Update. The Load operator (FastLoad) does not support loading views, but the Update operator (MultiLoad) does.

I believe it's a bug in the job compiler. The environment variable job parameter should be represented in the osh as [$DSCAPIOP_PARAMNAME] instead of #DSCAPIOP_PARAMNAME#. Try getting the patch for Ecase 104775.

Yes, you can specify multiple statements separated by semicolon characters. Just don't use a line feed.

0A is the ASCII hex value for a line feed character which the Teradata CLIv2 does not like. Try taking out any carriage returns in your statement. Does your DELETE statement span multiple lines? If you did a cut-and-paste from somewhere, make sure any terminating line feed was not included in the cu...