DSXchange

I would suggest you search your server system for files matching the pattern "*7416.323000*" - perhaps the log file(s) are in a different location.

What do you mean by the example of "wrong date" - if it is an invalid string that fails conversion to date or a date constraint then the reject handling works as expected. I can't think of a case where a record wouldn't get inserted/updated into the table and also not go to the reject link.

Are the listener and other processes running? Does "telnet localhost" give you a DS login prompt (if yes, enter "UV" or a project name, and then "QUIT" once you get the TCL prompt)

In order to do what you are trying to do you need line terminators. It would be easiest to define just one column as a VarChar(). Then use the LEN() function in a transform stage to decide if it is a valid row and then parse the one big column into the separate components.

It has been a while since I did mainframe DS so I will have to pass on this one.

I migrated from one Oracle server to another Oracle one. In retrospect, I would recommend re-installing from scratch. IBM support has an internal document titled "Steps to Move the Metadata Repository to a Different Database" which pretty well covers the steps; I currently have a copy because I am o...

Is this an OS390 or a UNIX Parallel Extender job? If a PX job, use a reject link from the DB2 stage.

do you have constraints in the DB on that column? How about doing a simple test job with a single row and do an "update" on just that column using the derivation of "N" in a transform - has the error changed or gone away?

Is this user-defined SQL or automatically generated? What is the datatype for DIRECTLY_TIPPED?

Perhaps it is another field - often it is a non-nullable field that isn't included in the INSERT statement.

No, the only Windows DataStage version that allows Parallel jobs to be run is 7.5x2

I think it does a "DELETE *" and if you want an efficient truncate you will need to execute a "import from /dev/null replace into..." command.

The MQ Connector doesn't use a "reject" link as would other stages. If you try to write to a full or non-existant queue the job will fail. If you try to read from a non-existant queue the job will also fail. The reason for the job failing will be in the log and will contain the MQ Error number.

Explicity enter one start each at 9am, 10am, 11am, etc.

vinodn - that requirement is like your company buying DataStage but demanding you use Informatica to load your data. The Oracle stage is used to connect to Oracle, the DB2 Stage is used to connect to DB2 and neither database is accessible from a Teredata stage.