DSXchange

if you do the commands: what is the screen output? Is your dssdev project displayed?

Even though this question has been posed and answered many times before and the search functionality will show many responses, I'll answer nonetheless - Your actual data length and your DataStage declared data lengths are not the same. In this case, ID is probably declared as a fixed CHAR() field of...

ketanshah123 - have you actually looked at the manual? In the functions appendix of the Parallel Job Developer's Guide there are CAST functions which you can use.

In that case, you should have no problems. The notation in the e-exponent form is just when you display the data.

Look at your options for DSAttachJob(), you can choose DSJ.ERRNONE to avoid an error message and can check yourself to see if the call was successful.

If you are reading this from a file or converting from a string, just declare your data type a float and you are good to go.

Have you explicitly sorted and partitioned BOTH left and right links? Sorted on the join keys and partitioned on just the main key? Add $APT_DUMP_SCORE to your job and examine the score to see if DataStage has inserted operators. Make the join output go straight to a PEEK or COPY stage with no outpu...

If the insert statements are indeed standard SQL then the approach might be to take the string, replace table/instance/column names as required, then pass that string to your Oracle stage and execute it directly. If it isn't standard SQL then it is not difficult to parse the string in a transform st...

Since your date format is YYYYMMDD you can use straightforward string/numeric checks for values greater than a less than, i.e.

Could you post a representative portion of one of the scripts?

Add debugging calls such as DSLogInfo() or DSLogWarn() in your routine to see if the values passed in are actually what you expect or to ensure that the routine is actually being called at runtime.

If you correctly define your character sets in source and target then you won't get conversion errors. Odds are you have incorrectly declared your NLS settings in the SQL server source stage.

5TB of source data transferred through DataStage with no modification or conversion can be done on a notebook. More complex transfers cannot.

We need other details.

Harini wrote:...subArgs={asc}}}(0),1: Fatal Error: Fork failed: Resource unavailable; try again...

It means that your OS couldn't start another process due to resource limitations, most likely your virtual memory is exhausted.

Can you post the derivation in the transform stage where you call the routine, perhaps it isn't doing what you expect.