DSXchange

Theoritically, the job should remove all the temp files created, once the job finished ok, but due the reason, the job is aborted, those files are left orphaned.

If the log is required for analysis, dsjob command has the options to retreive it. -log
-logsum
-logdetail
-lognewest

Optionally, it can be done by printing the log to a file from the DS Director from GUI.

What error?
Basically, Aggregator handles dfloat. If you get the integer value, you should be able to find the way to use aggregator.

If understand properly, does your source file has both data as well as the name of the target table to which it has to be loaded?

DS Metadata might match the table metadata, but check if the number of column given in the insert query is in equal number.

In turns its an hashed file. You can find under your project directory.

Hi Boppana, The filter option available in the Sequential stage is for pupose of executing some operating system command. As said you can use sort as in unix. You an simply 'cat' the file from the filter option and you can skip the filename option by giving /dev/null. Other than that what is the fil...

Project name is case sensitive. Do you got any error code?
Try to skip /NUA option.
What meena provides is for Export.

The pattern of filename might not parsed corectly. Check the value of parameter passed to each from the director log on second or third event. Check the value that been passed when you recompile and rerun.

If you want to produce a report of the rejected rows, you can either opt for the dsjob -linkinfor or DSLinkInfo.

Do you want to call this from datastage or you want to build the logic in datastage?

If a job failed abrubtly, there will be some event, which will be unlogged. Those unlogged evern will be bought out when doing reset, via "From Previous Log..." eveb. This is the reason for asking to reset when failed. Observe the pattern when it failed and when it was tryed to run with success. You...

Signal for segmentation violation may even occur, if Null been handled unexpectedly. Is there any Buildops or any external routines been used?

It is another similar function as what you tired. Julian data is represented in integer format. It has its own reference date. And the number of days been calculated since that reference data till the given date. The functions available to convert to integer - JulianDayFromDate(). Do the aggregation...

Yes.