DSXchange

Can I limit amount of rows when I run the job?

The "Limits" tab in the Director isn't available.

Either use a user with access to the LIMITS (developers) or put in a parameter and program logic to limit the number of rows processed.

Unfortunately row limits are only available to server jobs, not parallel jobs. There is no easy way to do it in parallel jobs.

The sequential file stage has an optional option called "Read First Rows" and you can set this and pass a number in as a job parameter. There is no easy way to turn it off other then removing the option and recompiling the job.

Database stages may support a row limit in a where clause, such as a row number statement in Oracle or a TOP statement in SQL Server. Once again they may be difficult to dynamically turn on and off.

what is your input file as, if its a database file then you can surely limit the rows according to your requirement... if your imput is any of the databases then use these following querys to limit the rows and if your input is a dataset or a sequential file then use a transformer and RUN it in sequential mode later on go to constraints and use @inrownum<'your desired number of rows'...that should work.

-DB2 select * from table fetch first 10 rows only
-Informix select first 10 * from table
-Microsoft SQL Server and Access select top 10 * from table
-MySQL and PostgreSQL select * from table limit 10
-Oracle 8i select * from (select * from table) where rownum <= 10
-Teradata : select * from table sample 10

10x

Thank a lot for your promts.
If you have any other idea please tell me

Regards,
Anka

But the queries retrieve random rows and I am not sure if they pick the first ten...incase if you want the first ten.

As long as you do not want the data ordered by and not go too much indepth with the row migration, they must be much accurate. Anyhow it must be fine to allow a sample rows flow through.

I don't need retrive the first rows.

I just want to limit the source rows.

Well, Firstly whats your requirement? then whats your input ?as I said If your source is a database then do a group by and use this funtion then you will get the first 10 rows, and if your souce is any seq file then you have to necessarily use a Xformer and run in sequencial mode so that you can stop the inrows to what ever your req is....

My Bad.... I mean ORDER BY