DSXchange

Simple, but works as required. Transformer: COL1 --> is your input with two fields separated by ',' Declare Below Stage Variables: svIndex - Index(DSLink1.COL1,',',1) This will find the position of ',' svLength - Len(DSLink1.COL1) this is to find the length of your column svField1 - Left(DSLink1.COL...

Hi, Thought of sharing a solution I have built to replace a sub-string in the string (similar to EReplace), but it is not a function. (it will work only for the single occurance - I will post the solution for multiple occurance as well) Input: I have ID and DESCRIPTION fields and Description has inv...

If these values are dynamic, better use lookup. Keep these values in either a file or table (depends on your access) then you won't have hassle of changing the datastage job every time these values change.

If you need to delete all the records only when all the records are inserted then:
First insert all the records into table2 (create this temp table as before sql statement)
have the after sql statement to truncate table1 and rename table2 to table1

From the error message, it seems the multi-row fetch/insert is above the limit allowed. As per the below link, 32767 rows are allowed. http://www-01.ibm.com/support/knowledgecenter/?lang=en#!/SSEPEK_9.0.0/com.ibm.db2z9.doc.codes/src/tpc/n246.dita Try loading 32760 rows and see if you can load then t...

Since I can not use the Oracle reject as reference link to Lookup, I have used a Copy stage. Though the input link to Copy stage is failed, output produced zero records..Is there any way to avoid this..

I have a job which extract the records from Oracle table and inserts into another oracle table. All the records successfully inserted should be deleted from the input table. For this I have below design: OC Extract Copy ---> one output is going to OC_insert and another is for Lookup OC Insert --> I ...

Right Click on the Palette bar, then Customize Palette, then right click on Current Palette groups, Select Add top-level group. It will create a new group. You can drag and drop required stages into this.

Try declaring your input column as SQLtype - LongVarChar, Extented - Unicode and don't mention the length

This article from IBM explains Oracle Connector (you can read specifically - Parallelism and partitioning part for your issue) very usefull and in-detail explanation of different aspects.

http://www.ibm.com/developerworks/data/ ... onnector1/

from the error message, it looks like the validation is not able to find the HeaderMsgType and ProviderType. If these are two external references (meaning, if your original xsd is referring to another xsd) make sure you import these xsds as well. Below is the example: --This is just an element <xsd:...

You can try below:
InputField:Str('0', 4 - len(InputField))

InputField -- is your input field
4 - length if your output column after padding the leading zeroes

--this will not work if you have declared output field as numeric.

Though your reference link has duplicates, unless you specify "multiple rows returned from link" for the reference link, you should not get the duplicates. If your main link has duplicates, then it is possible to get the duplicates.

Hi, Sorry if this is a repetitive post, I have searched and couldn't find any thing. I have a situation where I am extracting data from Oracle table with more than 200 columns and loading into an another Oracle table with same structure. I need to select all the columns. I have to do some transforma...

Thanks Craig..But in my case there are 90% records matching and only 10% of them are not matching (average volume in the input is around 100K). I can have a filter stage to get only the null value records, but in this case I will be checked the null value for an extrat 90% of the records. This job i...