DSXchange

There's no additional information logged when you Reset the aborted job? Specifically in a 'From previous run...' log entry that the reset would add. I'd also be curious what exact Oracle client version you have installed on your DataStage server? Some are buggy and can cause these odd intermittent ...

If the source data is larger than the target column size, you should handle that in the job, either by logically rejecting it via a size check or substringing the data to fit. Oracle will abort if you don't, as you've seen.

Any such limit would be on the Oracle side.

Well... it is but if the only difference in the 'configuration' is the number of nodes, there wouldn't be any issues.

'Z' means the leading zeroes are suppressed.

You could try incorporating some kind of substring check based on result length and key value in the transformer to weed out the false positives... maybe. Sorta.

Could have told you that while that index observation was an issue it certainly wasn't the issue. No-one can help you with this except you and yours, especially your DBAs. Resource busy means the table is in use. Period. Unless you are the victim of some bizarre Oracle bug, as hard as that is to bel...

It is included in the 8.x release now but (since it was a chargeable item) I believe you have to ask for it for the 7.x release.

ORA-00054: resource busy and acquire with NOWAIT specified DBA's have confirmed no one else is loading these tables/accessing any object in the schema Short answer is the DBAs are incorrect, as the error shows. Someone or something has the table in question 'in use' such that your operation cannot ...

PRODUCT CATEGORY *********** *************** PROD ID CATG_ID CATG_ID CATG_NAME PROD ID 1 100 100 ABC 1 2 101 101 PQR 2 3 102 103 RTS 4 There you go, all better. Code tags, people, code tags preserve white space.

rumu wrote:I heard that in such scenerio 2nd job shud abort

You heard wrong.

To what purpose? Wondering if perhaps you really mean 'read only', hence the question.

#2 doesn't make sense but #1 does. Do you need to know which columns are null or just the fact that at least one column is null?

That would a dynamic (type 30) hashed file and how you remove a hashed file depends on how it was created. A pathed hashed file where you specify a directory for it to be created in is created by the operating system command mkdbfile and can be removed with a 'rm -r <dir_name>' to remove all of the ...

I have no idea what you mean by "run c code". The question is pretty simple - did you create and compile a small C program using the compiler options you posted?