DSXchange

You may need to add a link from the execute command to the notification activity followed by terminator stage to abort the process.

Trigger option of the execute command should be custom conditional and define your condition there in case of unexpected output.

As said by Ray, Data viewers sometimes shows a special character but the data inserts into the table may not have that character.

Do you get any errors because of this special character when you insert the data into the table ?

The routine should be present inside the Routines folder in the repository. And this looks to be a server routine and so for the routine to work, the Routine transforms also needs to present. The routines are called inside the transform, which in turn is used in the Basic transformer stage. Am not q...

Instead of table, can you load the data to the flat file and check if you have any special/ chinese characters .

This may give you some idea Wait for File Activity -> Transfer / FTP file to DS directory -> execute the jobs -> Delete the file in Local (The same process continues for the next file. If you dont need to delete the file then probably you can timestamp the file as you are done executing the jobs. Th...

Alternatively, you can use Complex Flat file that supports multiple records. Which system does generate this file ?

Yes you are right prasson. If its duplicate then the data rejects are obvious.

Instead of doing this in the Transformer you can extract only those 50 characters from the database.

Eg : SELECT SUBSTR(NAME,1,10) FROM CUSTOMER

Does your NLS Char set in Datastage support the Chinese character set ? What is the error that you get when you use column1[1,50] ?

You can probably check the wc-1 (word count) of the file and use this condition in Custom Conditional -> Trigger tab

convert(@FM,"", exec_filesize_isst.$CommandOutput) > 0

This means if the file size is greater than 0 it would trigger the subsequent job.

Column Import reads the multiple column file and generates a single row out of it. You can further use transformer to split the single row into multiple columns.

Column Export will read a single input and splits it to multiple output columns.

Is the key unique on both the links ?

May be the user you connect to Unix may not have privilege to execute dsjob command. You may need to check the user credential.

What does the routine does ? It just captures the log details ?

After pivotting the data, you can add a transformer and check the outcome of Column6. If Column6='' Then 'null' else column6. I am not sure if this is what you are expecting. Pivot Enterprise Edition in 8.5 has checkboxes to select the Group by columns and columns that you may want to Pivot. If you ...