DSXchange

Please provide more details...are you reading xml, trying to write xml...which stages are you using, what are your sources, etc. etc.

Thx.

Ernie

Most likely it is perfectly reading them......but they are "in the job" somewhere. It "appears" that they aren't being read until the timeout ends, but what is really happening is that they are stuck in a buffer somewhere, and when the timeout ends, the job is signaled to shut do...

Transaction semantics and messages are two separate things.....quite often they are indeed 1:1, but sometimes they are 1:many, many:1 and even many:many --- it depends on the payload of the message. Transaction semantics is a whole separate dialog in itself.....as for messages, however, and to your ...

...and be careful...you may not get a whole lot larger than what you have already.....at about 500M things hit limits in the current Stage.

How close are you to moving onto 8.x? The metadata workbench gives you all of this and far more....and some sites are starting just with reviewing their metadata...loading up their 7.x jobs into v8 and then using Workbench against their 7.x designs while they prepare for the actual migration of the ...

You have elements. They just don't have "lowest level values" in them. You need two links.....one for the ReportDefinition path, and one for the ReportBody path... Put a column in your link, with your attributes, for the container element (ColDefData and CellData respectively). Give them &...

Hard to say, but they might be related....check in the DS Administrator..for this project, make sure that you do NOT have "collect operational metadata" turned on........

Use "ls". "pretend" it's unix.

JavaPack is an easy way to do it, assuming you are comfortable using HTTP within Java and then can spend some time building an interface for JavaPack ....I've also heard of people using PERL for this as well from within DataStage, although I haven't done it.......I could be wrong, but I seem to reme...

ditto... but remove the d: ...that is likely the issue. The External Source thinks it is running in a unix environ....

You are experiencing the classic end-of-wave issue. The problem occurs in Parallel Jobs because the buffers generally can't be set small enough to force data thru the entire Job, and it will happen in Server Jobs (and Parallel Jobs) if use are using a blocking Stage such as Aggregator that requires ...

that is a namespace prefix. How did you import your metadata...if you imported via the xml metadata importer, and you "loaded" your column names from the "load" button, be sure also to use the "load" button also in the Transformation settings/namespace tabs to include t...

Sounds like the RT vs OR logging issue....... consider switching the flags inside of dsparams (has been discussed in the forum before...do some searches)

Ernie

I haven't seen any way to support the xml stages via schemas, but you can certainly explore the ability to pass xslt dynamically into the stage....not quite the dynamics of RCP that you are looking for, but it might accomplish what you require....the columns on the output link, though, will have to ...

And you only have one column in your job that reads the queue? A large varchar? ...and the Job is something like MQ to Transformer to Sequential?

Make sure it is that simple. Also make sure that your message is still in the queue.

Ernie