DSXchange

The Sequencer has no capability to feed parameters from a file into a job via the Job Activity stage functionality. However, if you write a function (which Tony has) which reads a file, parses the contents, and returns the values into fixed return codes, you can send those results into the job param...

You have an "improper datatype". You are doing something where the data cannot be used in the form it's presented. You are probably doing math on a string with letters.

This is a bad idea. Yes, it could work under specific circumstances, however, the caveats would be long and lengthy. The burden of simply importing and compiling jobs is not that great. Perhaps you could post why you are considering this route, because maybe you're doing things in a manner that make...

No, the Sequencer lacks any functionality to do this.

You should manage this via user groups. Let Unix level admin persons take a group out of access, because as DataStage has ZERO functionality in this regard. The closest you have is to use Admin and revoke all of the groups with access. Then, you have to go back and add them again when you're finished.

I suggested that you put the list of unused numbers into a hash file. So, have a job read the sequential text file of unused numbers. The hash file should be defined with two columns: <1> OUTROWNUM (check this as the key to the hash file) <2> UNUSED_NUMBER (this will be an integer) Now, in the trans...

There's lots of ways to feed the unused numbers in. One method is to put them into a hash file using a primary key column derived using @OUTROWNUM. In your transformation job, when it needs a surrogate key, simply does a lookup against this hashfile using @OUTROWNUM as the keyexpression and returns ...

Okay, you don't have a big table. How about a medium table? This creates a virtual big table: select rownum from (select rownum from mediumtable union all select rownum from mediumtable union all select rownum from mediumtable union all select rownum from mediumtable ) where rownum not in &#...

Sure but it's slow.

The filtering will allow you to selectively view the messages. You'll have to discriminate by dates only. The only other way is to manually get rid of messages.

An option is to programmatically extract all messages to a file, throw it into Excel, and parse from there.

Sure, just find a big table somewhere and select the rownumber from it. In Oracle, I'd do: select rownum from bigtable where rownum <= 10000000 This will return a list of numbers from 1 to 10000000. You could actually get the list of unused numbers in a SQL statement like this: select rownum from bi...

Only by working with job parameters can you achieve such a goal. You can put a job parameter as the password and supply the appropriate value at runtime. Environment variables work the same as parameters, you'll have to declare the envinronment variable as a job parameter. I don't believe a macro wo...

Under the DataStage Start button menu is Online Documentation. dsjob is extensively covered in the command line document.

Awe schucks, thanks for the gold stars. Actually, credit where credit is due. Steve Boyce handles all of the infrastructure and his talents include developing the underlying framework of scripts and process on which our ETL framework rests. You see a sampling of the skill he brings to our team.

Sure, check out mine:

http://www.dsxchange.com/viewtopic.php?t=85578