DSXchange

What stage are you using for this - the OraBulk stage? So that you are manually issuing the sqlldr command? Since the files cannot be appended to, you'll either need to rename them after job or perhaps pass in a job parameter to use when naming the files so they are unique, either to the file or to ...

prasannakumarkk wrote:Am not sure whether you checked this environment variable APT_TRANSFORM_COMPILE_OLD_NULL_HANDLING and replied.

Seems to me all that would do is preclude the use of SetNull() in the stage variable derivations. However, it can certainly be tested... more grist for the mill.

I don't see anything wrong with the derivation you posted, it should work exactly as you say you need it to. And if you are careful with spreading the derivation to the other columns and ensure you're using the correct input in the check to match the output column I don't see any reason why you woul...

If you are in 8.5 and still null is not handled by stage variable It's not the stage variables are "not handling" the nulls in 8.5, the logic being used isn't. Do us a favor a post ALL of your derivations for the columns in question. Seems like that should be 4 in your test job? I'd also ...

After the job completes, you check the row count down the link or the size of the file and send an email if it has been written to. A Sequence job is probably best to control this but it could be done via an After Job routine or batch file.

StringToDate() with the proper format string should get it to a date just fine. If you need it in a string in a certain format then use DateToString() after that. Those date format elements are all documented but "%m/%d/%yyyy" should work fine for you.

I seem to recall a SLEEP or NAP function or you could do the same thing at the command line via ExecSH. Do it before job and the job will "start" but not process anything on the canvas until the sleep ends.

Not sure how often U checks in here.

Tell us - is this a new installation and has this never worked or was it working before and stopped working? And is this just a problem for you or for everyone?

And? Pass into the job whatever you need to use to update the record. If that means the "unique id" then that becomes the job parameter. I would still suggest a small Server job if that's all it needs to do, you can build one with just a Transformer and a DB2 target that sends one row to t...

Have you tried passing your "n" times in as a job parameter? I have no idea if that's allowed in the stage but for it to be possible I would imagine that's about the only way you could manage it.

some how one of the record is getting rejected You need to understand that hashed file columns are positional and you are asking for nothing but trouble if you leave out columns or specify them in different orders unless you really know what you are doing. And it's hard to help with an insert issue...

Detail for us how your FileWatcher is configured and what your trigger expressions are that come from it.

I know you're not looking for an answer of this nature but in my mind this is where a decent Enterprise Scheduling tool would come in handy.

One way: viewtopic.php?t=127425

Well... if you're having DNS issues, those should be looked into and resolved as well. By someone else, thankfully.