DSXchange

Umm... You could post-process the generated XML code to remove all carriage returns other than those following a closing tag (or remove them all then place one after each closing tag). Don't know how to do what you want with the XML Writer stage, but the above would be quite straightforward to do in...

If the two tables/views are in the same database, you can also use the passive stage that extracts them.

When building the SQL, add the join condition to the Selection criteria, and inspect the generated SQL. If it's not quite right, change to user-defined SQL and edit what was generated. :D

Did you even stop to ponder my solution? I have many years experience with this product, as my more than 2000 posts here suggest. This solution will work irrespective of source and target.

See also http://www.dsxchange.com/viewtopic.php?t=86415

That makes sense, given that init (pid 1) runs as root and starts things in the auto-start directory (/etc/rc2.d or wherever). However, it is my understanding that dsadm can use the -admin option to stop and start services using the uv command. I can't test this just now - perhaps someone else can?

Multi-row result set is a property of the ODBC and UV stage types only.
See help topic "Defining Multirow Lookup for Reference Inputs" in the Help from Designer or Manager (dsdesign.chm).

How this is done varies between parallel and server jobs. The piece you've quoted from the manual relates to server jobs, for which Vince provided the clarification that it is the Transformer stage in which rejected rows are identified and redirected. Unfortunately I do not have access to PX at the ...

You might have more luck posting this on the Parallel Extender forum.

Probably the easiest is to run the data into a UV table, with the "Page" column defined as multi-valued, normalized on the Page column.

Then use either the synthetic key @ASSOC_ROW to limit the values that you retrieve from the dynamically normalized table.

Read the installation instructions again. The only possible user ID (apart from root) that you can use to install DataStage is dsadm. Were you using this ID? There is a separate chapter in the Install and Upgrade Guide (which installs with all the other manuals with your client software) about how t...

checking for a successful DSSetParam call to ensure it was ok for each parameter you set in each job. No less important, if you "roll your own", is checking that DSGetParamInfo calls were successful. Build in insurance against future developers, who obviously know/care less than you (always a safe ...

Neater and cleaner: do it in a job sequence, or in a "roll your own" job control routine. This can pass the date to the main job (parallel or server) as a parameter value. Another way is to extract the date into a text file, as you indicated, then load an environment variable with the date. CREATEDA...

LK_cleanup_pshm is the module called by uvdlockd (dsdlockd in later releases) that cleans up locks held by the defunct process that owns the printer shared memory segment, as part of reclaiming that process's resources. It can "fail" if it can't clean up the locks because they're shared locks also h...

Was it a large difference? What are your measures? Rows/second is not a reliable metric. Can you provide elapsed times? Total volume of data? Did you take any steps to eliminate the effects of cache? (For example, the second run will benefit greatly from operating system and disk controller cache). ...

There's no such utility in the product (though you could create a DataStage job to do it). Many folks have created such a utility; a search on the forum should find a few of them. It can be done from the TCL shell (or from the Administrator client Command window). DIVERT.OUT ON &UFD& dslog.txt SELEC...

Welcome aboard! This topic has been covered in the past, so why not begin by undertaking a search of the forum (see Search in the menu bar). Likely useful topics include CDC, "change data detection", "detection" and CRC32. You might be surprised to find that others have already been down this path! ...