DSXchange

Look into sed, Craig. Something like

You can perform this command on the files, either before concatenating them or after.
As for the speed, you will have to test it out. Sed is pretty fast. But dont know with 51 x 2M records.

Or instead of sending to aggregator, load it to hashed file keyed on ACCT_NO. This will automatically retain the last row.

Das wrote:and i the input tab file name as input link name and in output tab output ink name.

I think this is where he is going wrong. Giving different names while creating and reading the hashed file.

The hashed file name should be the same while creating and reading the hashed file.

There is IsNull() function. You can use that to identify null values and constraint them or pass them down a reject link. If your source is a flat file you might also want to test for zero length.

vnspn wrote:The incoming data is already sorted on the column that I want it to be aggregated. The source is from a flat file.

Plus, the sorting is to be done and specified on the grouping key and not on columns that are aggregated.

Your specification is wrong. Provide the following initial_date:date = date_from_string [%mm/%dd/%yyyy] (initial_date) And if your loading to a flat file, in the sequential file stage, go to "Format" tab, under "Type Defaults", go to the "Date" folder and enter %mm-%dd-%yyyy for "Format Stri...

Are able to get a result of "greater than 0" if you do a count(*) from sql plus? If yes then dont worry about it. Run the job and it should be fine. I think it has no bind variable and hence not able to fulfil the where condition resulting in zero data.

So what your saying is, that all the 100 table definitions are identical?
That does not make sense. Can you be clear on your requirement.

Try compiling with +Z option.

You date is in the format mm/dd/yyyy and hence you need to give that format to the date_from_string() function. Then while loading it to a file you can specify change the default date format to mm-dd-yyyy.

Please search on keywords "General repository interface 'other error'". Do an exact search. This has been covered before, a lot.

Yes. Search for "remote shell", it has been covered before.

Maybe remote shell Or install samba, this way windoze will become transparent and you can see/execute files on windoze.

I think you mean, while in vi, save it as follows, right smrutiranjannayak?