DSXchange

Depends on what you mean by "limiting"....it is simply a way of identifying how the tool should "dynamically normalize" a hierarchy into a relational engine (which, when targeting an rdbms, is very likely part of the target model). The new XML Stage doesn't have the concept of a ...

I've never tried using that technique in the External Source stage... I would suggest you use list (ls) and send the list of filenames, and then change the check box to "url" inside of the xmlInput Stage (see the table of contents at www.dsrealtime.com for an entry on xml and how to define...

This is one of those "poorly designed" (imho) XML documents. Much better would have been a schema that actually has elements with the given names of <Empsal>, etc. [where you could then use the text() syntax directly]. In this case, <Field> is a repeating node...with <FieldValue> a repeati...

It's important also that you outline what happens....0 rows? an error? nulls for those columns? As Ray noted, share your "xpath" for those columns (this is the syntax that you will find in the "Description" property of each column on your output link). Retrieval is dictated by th...

hard to say without "seeing" the input to step number 5...... send the output from that .../telex chunk to a flat file without any formatting (no quotes or delimiters). Look at it....the answer may be clear once you see it.

Ernie

There may be something else odd going on here also..... seems there might be a CRLF actually in your data. You might want to try using some ereplace functions in a transformer and changing all char(13) and char(10)'s [hex 0D's and hex 0A's] to blanks [char(20)] before writing to your target.

Ernie

Hard to say....but looking at the 8.5 solution, one concern may have been the desire not to get boxed in. The 8.5 solution uses Information Manager, which is Eclipse based, and takes advantage of source code tooling that "supports the Eclipse Team" standard.... ...but there are many reason...

...of course, I'm talking about Jobs that actually "have" ISD stages in them......if you are deploying ISD Jobs that don't have ISD Stages (this is a supported pattern where ISD "kicks off" a Job that is not 'always on'), then naturally, they will have nothing running (as a requi...

It's just the ISD stages.......every other stage in the Job can be running elsewhere......Like Sequencers, those stages use various bits of technology (to talk to the ASBagent, etc.) that only exist at the conductor node.

Ernie

A separate config file is a good idea. I've seen VERY few ISD Jobs that really needed a grid configuration to sustain their transaction rates, and only one scenario that needed dramatic parallelism "within" the ISD Job (it was a huge "what if" scenario that performed huge joins w...

Excellent! I'll have to give it a try in 8.5, assuming the patch moved forward successfully.

Ernie

Which side is "driving" the transformation request? Is the portal the "trigger" (ie...when a user clicks on something)? If you have Information Services Director (ISD), you could create a web service that performs data access and transformation in a DS Job, and call that web serv...

...and as far as grid goes, WISD is pretty much abstracted [hidden] from it. What goes on "beyond the conductor node" (where the ISD stages run) is up to DataStage....

Ernie

good one Craig! ....and as long as we're counting, how many DS Jobs have you written in that time span..... ?

Happy Holidays!

Ernie

There was a site who had done this already...do a google search on biogrid australia.....you will find an IBM document that describes what they did .....last I heard it was password protected, as any such access should be, but there isn't anything in the technology that should prevent it.

Ernie