DSXchange

Thank you Craig I just had to go via conventional way of querying table to find distinct or sort(as you proposed) and found the problem area. Is valid sounds like an another fine alternative for these scenarios . However I feel Datastage should have the ability to throw the problem record in the log...

pavi wrote:which method did you use while importing metadata(table def)?try to import metadata from connector wozard or orchestrate table def.this will rule out any possibility of data type compatibility issues.

Yes , I am using connector wizard.

I think I did miss one fact here . My target reads on of these varchar columns as Number . Source datatype of a set of columns are varchar2 and target is decimal (number) . So when I read these columns I cast them to Decimal so that I do not have to do it in the transformation logic. and yes there a...

Hello all, I have two data source ( oracle connector ) connects and extracts these two combines and puts in dataset (simple funnal) I see the following error The OCI function OCIStmtFetch returned status -1. Error code: 1,722, Error message: ORA-01722: To further debug , I design a simple job where ...

This is now resolved using list option as Ray indicated above within the loop. Thanks Craig and Ray

--Sorry getting back to this one after some time now Ray I did test this . I took one job which calls a table (via parameter called table name). My design starts the loop with list option (comma delimited) and I gave two table names here in the list. This connects to a job activty stage where the pa...

So basically do we still need a script when we choose to do it via sequencer job ? There is one job that asks for source and target table name (parameters) for each run , when we do it manually. When I call this job via sequencer and loop it - Do I still need a script to read the list of table names...

When I loop the sequencer,I will loop it x times but how can I change the table name (which is a parameter) on the fly to take one value (table name) at a time ?

So basically How do I say for loop 1 - Take the parameter as Table X , For loop 2 read the Table name as Table Y so on .

Hello All , I have x number of tables that I am currently loading via RCP . Since this is a direct load , I have to manually give source and target table (identical tables) names as parameters and run the job. I was wondering if there is a way to automate this by say making the job read the table na...

Ahh ..
1. I didn't look hard enough in the properties
2. I should have searched in the forum (never clicked I would find "Preserve type" answered in the forum.


Thanks Craig and Arnd . I tested the job and all that was needed was to set the property "preserve type" to true

Ok , I get it now , thanks for the clarification , String sort is allowed . However the maximum value as I mentioned above needs to be calculated only based on a non empty string (address data with numbers and alphabets like "240 abc" vs "empty" fields. I am still trying to find ...

Not sure I understand what you mean by "string sort" in this context, here we are comparing empty string with data (which is a combination of numbers and alphabets) . In this scenario its always "empty string" vs some data (address data) . There is a preserve partitioning drop do...

Thank you craig - I think I was referring to something like seq() function that converts the first char into (i believe) Ascii , I am aware of the datatype conversion function you mentioned above, but as you mentioned since the aggregater is only giving out "23" between "23 xyz" ...

I want to try some thing here - and this raises another question - is there a way in datastage to convert char data (varchar datatype) into numeric - find the max of data (as it would be numeric now ) and then reverse it to characters (varchar datatype)

Hello Gurus , My design includes a seq file stage - an aggregator - seq_file_output file.. My input looks like ActNo , posno , phone_1, phone_2 , address_1 , address_2 1 , 1 , 1234 , , 12 abc , 1 , 2 , , 4567 , , 23 xyz 11 , 1 , 6789 , , 15 mef , Output should be (idea is to merge based on key colum...