DSXchange

DB2 plugin works differently than Oracle or ODBC stages. If the datatype is DATE, the plugin will silently do an ICONV coming and OCONV going. Change the metadata in the job to TIMESTAMP and see if your problem goes away. In the future make sure you always change your metadata. Otherwise, just put O...

You're in for difficult times. You'll have to avoid any Windoze functionality when developing, therefore, you'll have to parameterize all directory paths including the slashes so that you can correctly move to the unix environment. You'll also have to parameterize all calls to operating system comma...

Data is written in blocks of characters (4K, 8K, etc). There is no concept of a "row" in a file, therefore, simultaneous writers will interleave data in the file according to the block size. Also, you'll have to deal with locking mechanisms the operating system may impose.

If it becomes an epidemic, we'll deal with it. Otherwise, I'm not too worried.

We should mark John's posts for new members to review so that posts can reach the quality and attention to thought that he gives us. World take note!!! Ray's given you the slickest answer. Use a job parameter to contain your custom SQL generated by your job control. You can always add extra output c...

Welcome aboard! The easiest method is to take a date, say 2006-07-14, add one to the month and set the day to -01, thus 2006-08-01. Then, subtract one from the date. To do this, write a small DS Function. Here's a good start for you, it assumes that Arg1 contains a YYYY-MM-DD date: CurrentDate = Arg...

Link collectors are okay as long as the data is balanced on the links. If you have two links going into a collector, but only expect data to come down one link most of the time, you'll probably encounter timeout issues on the other link. You want to make sure some data goes to each link periodically...

We see all types of people here. Don't let it bother you. Even we've received nasty messages. I'm still trying to "climb off my high horse" over answers I've given in the past.

Here's a good chance for you to experiment. Copy your job and change it to write to another Sequential file instead of a database. After you run it you'll find that it is 1000's X faster. It's funny you assume that DataStage is the slow part. You need to understand that in data processing reading/wr...

I would go back to Ascential/IBM and demand a patch. They've just recommended a work-around that defies supported use! The Hashed file stage is for DataStage hashed files, not Universe. Just because DataStage is a code branch of a Universe release doesn't insure 100% compatibility. What happens when...

Add a CALL DSLogInfo line to output the results of your command to the job log so that you can see the value returned. You'll probably see your problem.

You can't move projects, so you'll have to create new projects where you wish and then export/import the contents of the old projects.

8.something, I'm not totally clear which version of 8, it changes. You'll need to visit the Ascential/IBM site and look at what features are scheduled for which verson. They've announced that there will be many scheduled releases of 8, starting with 8.0 and then 8.1, and so on.

There is no integration with any tools, Clearcase is fated for release 8.?

The poster still doesn't understand Ray's question. 10 million rows of one column is vastly different than 10 million rows of 100 columns. When attempting to tune the parallel jobstream you have to balance many factors: hardware characteristics, data characteristics, data volume. One single row of d...