DSXchange

No, I meant that was something you would have to do if you took the $PROJDEF route. For a Sequence job solution, your jobs would be unchanged but your entire job control methodology would need to be rewritten.

Sure, that's exactly how it is done. Using a 'find' command with the 'older than X days' option hooked to an 'rm' command is the way to go on a UNIX server. Can't say for Windows unless you have the MKS Toolkit (or equivalent) installed.

No. What Tony described is one of the many things Sequence jobs bring to the table - automation of parameter assigned being one such. What are you using for 'job control'? You'll need to find a methodology that is compatible with that. If I get what you mean my job control...then I run the jobs usi...

I would suggest you only clear it when no jobs are running in the project to avoid any 'issues'. Or do as most people do and every day clear it of anything older than X days, in essence keeping a rolling X days in there.

No. What Tony described is one of the many things Sequence jobs bring to the table - automation of parameter assigned being one such.

What are you using for 'job control'? You'll need to find a methodology that is compatible with that.

By building the appropriate 'command' for it to run - and that would typically mean 'script'. How are your scripting skills?

The User Variables and Job Activity stages are available in Sequence jobs... are you using them to control your processing?

Or... if you really wanted to go that route... store them in a file. 'Encrypt' the file if you like. You would then need to have some home-brewed Job Control in place to read (and possible decrypt) that information and dynamically set those values when starting the jobs. Many of us have such a beast...

But now, we are doing the same jobs in Datastage,we use HashedFiles as lookups.But the functionality of HashedFiles is if it finds 2 identical rows,it will retain the last row. Yes - it will 'retain' the last record for any key group sent to it, that's the nature of the destructive overwrite it doe...

You can also use the SQL Builder or Fully Generated option in the OCI stage. If outer joins are mysterious to you, it will walk you through the process.

Short answers:

1) Yes, any kind of load actually.

2) Yes, and the methodology is documented in the manuals.

As noted, you'll find many, many (many) conversations on those topics here.

:D You're overthinking it. Somewhere on the IBM site will be the full release history, but from what I recall the 7.x releases went: 7.0 7.0.1 7.1 7.5 7.5.1 7.5.1A 7.5.2 Something like that, anywho. Unofficial names are things like 'Twister', 'Viper', 'Sorcerer'... I think Vincent has them all in hi...

Which 'error' are you still facing? That you can't restart DataStage or you can't import metadata? You've posted two different methods - ODBC and native OCI - that you are attempting and which are (somewhat) mutually exclusive. Meaning, if you have OCI working why bother setting up ODBC? Both can ta...

No, there was a '7.5' first before any of the 'dot' releases.

Glad you were able to work this out. It can be hard to do this 'remotely', about all we can do is nudge you in the right direction.

I'd ask you to mark this as 'Resolved' but... it's not your thread.