DSXchange

What does what mean, the '3'? Did you search the forums for JOBTYPEIND, that will turn up the list of values.

1) How often do you need to load these files? As they arrive or once at the end of the day, or ?

2) Can depend on the answer to #1. A constraint can skip the first X input rows, if needed, but you may have metadata problems reading the file.

I'm guessing you haven't read the XML Best Practices document? Chunks are snippets of XML. They don't stand on their own, you generate 'chunks' of XML as separate, logical portions of the XML target and then 'assemble' them at the end of your process into the complete XML schema. We load them into h...

Ok, I can see the issue if the encrypted value includes those characters. However, it's one thing to handle the value as encrypted within DataStage and another thing entirely to pass it still encrypted to your shell script - that's not going to work unless your script knows how to decrypt it before ...

Describe in detail the changes you made that took you from working to not working. Where did the "<>" characters come from, they are not a normal part of the process as far as I know.

We stick with 0 as the transaction size for the same reasons Ray noted and why I consider that an 'advantage' - we get an "all or nothing" load. It greatly simplifies the restart after failure, just rerun the job with no recovery / restart logic needed once the problem has been corrected. Ray, you'r...

Talk to a SysAdmin or someone 'shell savvy' there, ask them about scripting attachments with sendmail or mailx. It will typically include the use of uuencode unless you have a client that supports attachments directly, like say 'blat'.

Use that with the DSGetLinkInfo function after the fact. For the aggregator, how about sending a column to it with a value of '1' and summing it? No grouping, just summing.

yaminids wrote:Can we use the ORAOCI BULK stage if DataStage (Linux) and database (AIX) are on two different servers?

Sure.

Are you using insert rows with out clear option as the update action in the OCI stage.That is the reason you are commiting after the end of loading all records. I doubt that. And that's certainly not the only reason to commit at the end of the load. If you are commiting after writing all records an...

As per the above error, my address sequencer job was in compiled state. No - the error says it was not in the 'right' state. And it defines 'right' as compiled and not already running. And the typical reason for that error in a Sequence job is that you tried to run a job that is already running. Sh...

Why not bulk load then?

Two jobs - first land the data from the ODBC source. Second, either script a sqlldr session or use the ORAOCI Bulk stage.

poorna_76 wrote:even though i see a space in the dataset, datastage does not recognise this a EmptyString.

A little lost. A space? If there's a space then you don't have an empty anything.

You can if they are defined in the project and then passed in as job parameters. You would need to use the function noted for environment variables not defined / passed in.