DSXchange

Why don't people read the messages?!! From Source_data (on partition #0 as it happens) a row has been read that contains the characters "NULL" in positions 123-126. Presumably this is one of the fields defined in the record schema. This field needs to have its Null Field Value property set to "NULL"...

What happened that this suddenly became resolved? Most recently we had

skp wrote:I am implementing the same as specified, but getting 211 of the first Variable only.

Now the thread is marked as resolved!

The Oracle Enterprise stage with Write Mode set to "Write" uses sqlldr anyway. As Arnd notes the Oracle Enterprise stage does have the ability to determine how the table is partitioned, and to use that information to load the partitions separately in parallel. That *should* result is shorter overall...

In RetrieVe query language these equi-joins are accomplished using TRANS() functions, either in the file dictionary or in EVAL expressions in the query command line. Read about the TRANS function in the DataStage BASIC manual and search the forum for examples of use of TRANS() and EVAL (which can al...

I agree. Other than filtering by job name (for example not matching *.*)there is no means in Director to achieve what you appear to be asking. But what is it you really want? If all runs supply an invocation ID the "base job" will contain no entries. Also, you won't be aware that the log table for t...

You should be able to use a Format stage ("FFC operator") to convert to fixed-width format.

Use the system variable @DATE. This is set when the job starts and does not subsequently change its value. Far less efficient is the macro DSJobStartTimestamp, which has to execute a DSGetJobInfo() function (and perform a lookup in the Repository) every time it is called. Others may suggest calling ...

You have a Decimal field, not a VarChar field. (16,2) is not valid for VarChar. If, having converted this to a string there are leading zeroes, this is how DataStage does things. It displays leading (and trailing) zeroes to demonstrate that decimal precision and scale are being handled properly. Int...

You need some NullToValue() - or other null handling function - in the test expression portion of the If expression.

Welcome aboard. Your Field() functions are good. Show us how you are attempting the concatenation. What is should look like is: 1) Field(lk_AccInfo.AI_HomePhone,'-',1,1)---->211 (svA) 2) Field(lk_AccInfo.AI_HomePhone,'-',2,1)---->790 (svB) 3) Field (lk_AccInfo.AI_HomePhone,'-',3,1)---->2820 (svC) 4)...

Pretty much what the manual says. The Change Capture stage compares "before" and "after" data sets and identifies the changes. The Change Apply stage takes the "before" and "delta" and from them can re-generate the "after" data set. This might be useful if re-running or reconstruction, particularly ...

You will need to re-think the description and post the job design here. A Funnel stage can not be used to perform a lookup; the closest analogy to what it does is to perform a UNION ALL of the rows from its input links. Do you mean that a Funnel stage feeds the reference input link to a Lookup stage...

Not easily in server jobs. If the number of different structural possibilities is small you might contemplate using a separate output from the Transformer stage for each.

Welcome aboard.

Because the IsValid() function does not support a format string argument the answer is "no" - you must convert your string to the standard (default) format - or change your default format - before using the IsValid() function.