DSXchange

Well, you could do something like this: Job0 ---> Job7 } Job1 ---> Job7 } ---> Seq (All) ---> anything else Job2 ---> Job7 } Or course, Job7 will need to allow Multi Instance execution in case they try to run at the same time. Or at worst case you could run Job0 thru Job2 in a...

What stage are you using to access Oracle? The OCI stages will automatically do the TO_CHAR for you if you pick the Date or Timestamp formats and let the stage generate the DML for you.

Or perhaps I should say I'm having a problem displaying search results and I'm wondering if it's me or not.

I can search and get a list of results, but clicking on a topic gets me a 'The page cannot be displayed' error. Something going on with the site?

ogmios wrote:this timeout is supposed to be for clients but dsjob may be considered a client too.

Yes! This setting affects both the dsjob -wait command and the DSWaitForJob function.

If you don't care about the data in the file that exceeds your current record length, there is a check-box in the stage to 'Ignore row truncation'. That should stop it from aborting. Other than the columns that Ogmios mentions, you could also try setting your record length large and bringing it in a...

You shouldn't have to create a template, the product ships with a template ready for use by DSSendmail.

How can you use it? I think the bigger question is are you sure you want to use it in a Transformer? In a PX job? You really need to send an email for every row that goes through your job?

Way too many geeks here.

Check this post for the syntax you are looking for.

Scientific Notation.

Assuming that you are using one of the OCI stages to write to your target, you need to know that the OCI stages need Date or Timestamp data in a specific format. You can read the Oracle OCI Technical Bulletin that ships with the product for the gory details, but dates should be in this format: YYYY...

For starters, you can use the Num function. Check the online help for specifics on what it considers numeric.

All in all, you are much better off not using either of them. IMHO. Best to make explicit reference checks so you know which is which and split your output stream, doing seperate inserts and updates. And as ogmios notes, it is vital that the indexes on your target table support these mixed actions.

That's an interesting thought. In essence, build a 'sort key' - a single concatenated field strictly for the sort process to use that could be pitched after that. As you said, one would have to make sure the fields that make up the key are formatted appropriately. Certainly an option to keep in mind.

I've done that in the past, using two files and all. Nowadays, it's pretty cool to use the Filter option in the Sequential stage to do the sorting 'on the fly' and not need the second file. The output of the sort command goes to Standard Out and the stage picks that up and processes it just as if it...

Looking for a little syntax help for a Filter command I am using in a Sequential File stage... I am dropping a delimited file to disk in order to leverage the UNIX sort command rather than the Sort plugin as it can be orders of magnitude faster. It's fairly simply when sorting on character fields, a...