DSXchange

Lose the "E" for Iconv() - it specifies "European" format (day month year order). Your test value is in month day year order.

Oh yes you can! Just qualify the table name (DS_JOBS) with the schema name.

UniVerse/SQL (DataStage/SQL) is compliant with standards, so supports qualifying a table name with a schema name. Each project is a schema.

You can't just copy an SQL script into the VOC file; it doesn't work like that. The VOC entry must have "PA" in field #1 and the SQL statement must either all be on one line or use line continuation characters. The COPY command achieves none of these effects. What exactly are you attempting to achie...

Was the helper subroutine "Record" or "DSR_RECORD" ? This is important. DSR_RECORD is used by DataStage to get something out of the Repository, and if it fails it usually indicates a problem with the indexing in the Repository or the connection from the client. It looks like there was a partially co...

This would be better posted on a DB2 forum - it's set up entirely within DB2 Connect - DataStage at run time is simply another client application as far as DB2 is concerned. It's DB2 Connect that controls where the data come from or go to.

You can use the Administrator client - connect to the VERSION project (or whatever you called it) and use the Command window for that project. Or you can establish a telnet session to the DataStage server machine, get into the dssh shell in the VERSION project directory, and execute the command ther...

Does your stage include a non-well-formed reference to a job parameter (for example not enclosed in "#" characters)?

Welcome aboard! :D This one isn't your fault; it's a programming error in the client software. You may be able to get a fix from your vendor. The error is in programming using the grid control that is used to get the report from the UNIX machine and read it. Meanwhile you can use a text editor capab...

Tell us what the warning and error messages in the log tell you.

Reset the job in Director and tell us the detail of any "from previous run..." event that is logged.

In the "Hawk" release (not out yet) the second user will be able to open a locked job in read-only mode. Prior to Hawk, however, the user who owns the lock has opened a database object (the job design in the Repository) for editing. A record-level update lock therefore exists, so that there won't be...

It is not possible. Except for table definitions, each type of component must be uniquely named within the project.

Invent a systematic naming convention, or use separate "sandbox" projects.

What exact version of DataStage? In the various 7.x release there is an increasing number of compilation options to make job sequences record checkpoints so that it can be restartable from the first activity to fail.

Not enough information. What stage types are you using? What is the source? Are there any constraints in the source select statement that limits the row count to 5 crore? What is the "rows per transaction" setting? Do any error/warning messages from Oracle appear in the job log? Are you using OCI st...

This is happening before you even get connected to Oracle. "Failed to find instance" usually suggests a problem on the DataStage server, either with the environment variables for Oracle set in dsenv, or in the tnsnames.ora file (or the fact that this can't be found).

To expand on Vince's comment, you are simply wasting CPU cycles - splitting your stream into 105 then putting them all back together again in the Funnel stage. Why??? What precisely are the two inputs to the Lookup stage? Hint: wrap your diagram with Code tags, then use Preview until you get it righ...