DSXchange

If you are allowed to create a temporary table in Oracle, dump the dataset data into a table where the child oracle table resides and then push down the join operation as SQL between both tables, the temp table as outer joined table.

What I have done in the past is seperate out inserts and updates. This is more of a performance tuning issue and how to handle large amount of data in already populated table which should handle insert and updates. First I separate inserts and bulk load them to target and then via another job run th...

Sheema,

Are you inserting and updating data in target in 1 job?

If I were in your place, I would always run 2 separate jobs, one for parent and one for child and set dependency of the job that is supposed to run latter on successful completion of the former job.

This way we can mainting data consistency plus have more control on the end to end execution

I am not sure if I completely understood the question, but have you tried to read the date/time column from source as CHARACTER. Try to see if you get complete value from source by putting the correct format of what you want.

Example: TO_CHAR(MXTR_LASTUP_Y, 'YYYY/MM/DD')

Also, do you have to absolutely use the Before Sub Routine for this SQL? If all you want is to run a SQL before the actual job, one can also design that as a separate job only for that SQL and then once the job completes successfully, call the actual job (the one after the 'before sub routine). Does...

It's not surprising if a Microsoft DBA doesn't know the information on UNIX because most of the time they do their DBA activity using GUI interface. OK, coming to the point, there are several ODBC drivers in market that one can install and connect to SQL Server and run queries. PERL is also an optio...

Extending to what Aruna said, try to decode the value of the column that is creating issue. See what the SEQ() / CHAR() value is and then make sure they are indeed spaces or low values

I am assuming the Stored Procedure doesn't take any input parameter. If that is the case then Open the ODBC stage and click on OUTPUT > GENEARAL > STORED PROCEDURE. Select the stored procedure using the imported definition from Manager. First run it successfull and dump data in a txt file or peek it...

Are there any job in that sequencer?

I beleive this error only comes when you have a Terminator activity stage without a Job Activity stage.

Set the option "Abort without sending STOP requests" on the Termination activity stage. You won't get the error any more.

Good Luck

If you have the ability to create a staging table in your database, create the hash file structure as a table and then use the source and the hashed Table as a SQL. Basically rather than using hash file use the data in the newly created table. This way one can PUSH the SQL to database with much fast...

If you have the ability to create a staging table in your database, create the hash file structure as a table and then use the source and the hashed Table as a SQL. Basically rather than using hash file use the data in the newly created table. This way one can PUSH the SQL to database with much fast...

Design 2 step process. (1) Step 1: Download the file as "Binary" as only 1 column for the whole row. For example if everyrow is 800 characters long in total in transformer/ or FT plug in stage define only 1 column like getdate=CHAR(800) rather than defining each and every column individual...

Expanding on what Chulett mentioned here is what I do in these situations While creating HEADER FILE I add a column called "Join_Column" and in populate the column with value "X". I create the same column in Trailer and populate it with Value 'X'. Then we can join the two feeds u...

Use ODBC Stage as you suggested in your email. Source > ODBC Stage (Server Job) From Source, pull the column name in the same order as mentioned in the INPUT PARAMETER and via a transformer drag all the columns to an ODBC Stage. In one of the TAB in ODBC Stage Use "Server Procedure", impor...