DSXchange

For a small finite number of known table names multiple output links from a Filter or Transformer stage would serve as a solution - one link going to each of the known table names. A dynamic solution would be more difficult. Perhaps use a server container (or server job) with a Folder stage to creat...

Try NUMERIC(19,4) for mapping MONEY, and REAL or FLOAT(6) for mapping REAL.

You cannot perform assignments in an expression. They end up being treated as Boolean expressions. So, yes, you need If IsNull(DSLink166.GEO1) Then -911 Else DSLink166.MD1 as the derivation of output column MD1. Or you could use If IsNotNull(DSLink166.GEO_1) Then DSLink166.MD1 Else -911 which is wha...

Use a reject link from the Oracle connector. Perform a difference comparison between those keys that the full set that was processed (to acquire those that were successfully inserted) and use those keys to update the source table. Or perform an inner join between the source where SRC_SENT_TIME is nu...

Good news! nsv2143 is running 8.5 FP1

No I'm not. The transport blocks are used when player processes need to communicate with each other, for example during re-partitioning. The ones I'm talking about govern pipeline parallelism, and have names beginning with "APT_BUFFER_". Values for some of these, such as buffering policy, ...

This sounds like a question for your Oracle DBA.

If the "Data which exceeds the maximum length got truncated" message occurred in the data browser (view data), that's a limit in the data browser only. You should ignore it.

If the lookup fails, then the returned value is NULL. But you have tested it with IsNotNull(), which means that successful lookups will return -911 and failed lookups will return the returned value (i.e. NULL). You needed to test using IsNull() or to reverse your logic: If IsNotNull(GEO_1) Then MD_i...

What happens if you add 0 to the value? As a general rule DataStage BASIC returns the shortest possible string as the result of an arithmetic operation.

I guess you could encapsulate the SP in a script that included both the exec/call and the commit.

Don't really know, but it's probably worth re-synchronizing the project.

SURA wrote:Consider the data type.

There are no genuine data types in server jobs.

Welcome aboard. The first question is "why?" The values are identical.

You can probably apply the Fmt() function to achieve what you require - read about this function in on-line Help or in the DataStage BASIC manual.

What stage type writes data onto the reference input link to the Lookup stage Lookup_103? What is the format of the ID column on that source, and what is the format of the ID column (or whatever column you have mapped to ID in the Lookup stage) on the stream input? Finally, what kind of Lookup (equa...