DSXchange

These aren't normal 'phantom style' errors - as a matter of fact they don't seem to be errors at all. I would think that you are using a routine or some piece of custom code that has debugging statements left in it - PRINT statements, for example - that are 'logging' information when the routine is ...

If commit level is set more than 1 then unless the number of records inserted is more than the commit level the records will not be commited. So in your case this may be the problem. What is the commit level set other than 1 ? I'm sorry, but unless this is in regard to a specific bug with the DB2 E...

No, that <name> must match the other [name]... confusing, I know.

And no, there's no reason to restart DataStage.

I would suggest you not modify any of those example entries, but instead copy them and create new ones. That and name them approproately because the name in square brackets is what you will use to refer to the connection in the ODBC stage.

Exactly... an empty trigger file, typically, that is only sent after the main data file has completely transferred. Otherwise you face exactly the situation you've described - the WFF triggers the moment it detects the file, finished or not.

Also referred to as a 'semaphore' file.

It will be checked against all output link constraints and so could go down any number of them.... including none.

Most people that post here as 'Server' peoples do so because their company has not licensed the PX product.

Ok... then you are ok now? My first post talks about handling a true TIMESTAMP field, do you need more information or are there still doubts? As noted, you do not want to drop the microseconds.

Interesting... your other option, when in a situation like this, is to hover over the link first rather than simply click on it. You'll see the actual link in the lower status bar of your browser, or at least you will in IE - don't know about others like Firefox. Then just don't click on it if it tu...

Did you search for DSSetUserStatus? Did you find any posts like this one? There are more, but it explains the basic mechanism. In your job that writes the row count to a flat file, also use the routine mentioned in the post to push the Row Count to user status. Then you can pick it up automatically ...

You don't have a fixed width file. If you did, all records would have the exact same fixed width - hence the name.

Nope. Create the metadata first then use it. The act of 'loading' or dragging the metadata into the job itself is what creates the link the Usage Analysis tool needs.

Then a DataStage job to read the value from the source table and stash it somewhere. The Sequence should then be able to retrieve that value and make decisions from it. Search the forum for USERSTATUS or better yet DSSetUserStatus for one way to rather easily do the 'stashing' and retrieval of that ...

If or how on the Premium Membership? The memberships help support the site... all of us 'Premium Posters' donate our time but there's still plenty of costs associated with doing something like this. Every little bit helps keep the site alive and you'll also get access to exclusive content in the nea...

I am not a premium member and so unable to see your full reply. Can you Please post that again so that i can see it. No, I'm afraid not. Sign up! Pennies a day, ya know. And i used to think that Array size is the no. of records that DS will fetch from a DB and Transaction Size is the no. of records...