DSXchange

Hmmm... a lakh is 10-Man which is 86A0x. Very clear... As Clarcombe mentioned, if your data is sorted then the Aggregator stage doesn't need to load and keep each & record in memory, it only needs to compute values until the next group level change. Using sorted input is the best and most effici...

Hell Saleguele, Doing a CLEAR.FILE &PH& will just clear the existing logfiles, it won't stop a job from continuing to run. It might be that your job is actually finished running (with an abort) but the Director doesn't know that. If you do a "ps -ef " grep {your-user}" do you see any process...

In order to do "array" type processing you could load your sequential as just one column. That gives a longish string, but you can parse that easily with the FIELD() function. In terms of execution performance, I think that the (ugly) (In.Column1+In.Column2+...)/((IF In.Column1#0)...

I think that the forum should also contain disasters/horror stories - it should show both sides. It shouldn't be a product-bashing area - that wouldn't happen too often, since the reasons for the failure are seldom in the product. I do agree that IBM/Ascential will use the "good" stories but that in...

You can create a loop in the Job Control portion of a job using DSBasic. There is a FOR ... NEXT construct, a WHILE / LOOP / REPEAT construct in different variations. I usually use the following type of syntax EQUATE True TO 1 EQUATE False TO 0 Finished = False LOOP UNTIL Finished = True JobHandle =...

Hello ruotahel, those "folders" in UNIX are actually databases (hashed files) to DataStage. They should not be deleted from UNIX, they contain information used to generate jobs; you will get one of these directories per job that you have in the project. The only places in a project directory that ca...

DataStage will do the conversion for you automatically. If you have an input VARCHAR(10) column called MyString containing the value "1943.100" you can create an output column in a transform called MyNumber, you can do a derivation such as "In.Mystring*3.14159" and it would convert for you automatic...

I just re-read your post and see what you mean. Have you tried doing a "reset" in the Director and looking at the "from previous run" to see what the error might be, or looking into the &PH& entry for that run? If you have a programming error in a routine, or maybe even call a routine that d...

DataStage Server variable are not typed. The system will perform an implicit conversion depending on what you are doing, thus you can do: MyVar = '10000000' MyVar = MyVar[1,4] * 4 MyVar = MyVar - 3 : '22' MyVar -= 1 If you want to force a conversion to a string type for some reason, you can do a MyV...

Navenn,

in your TRANS1 transformer you are performing some mathematical derivation on your data, but some value is not a number (i.e. it contains a character) and thus Datastage uses a 0 instead of aborting. So, check your transform stage for derivations...

Hello Dan,

if the files are created within a job and people also use the column definition metadata, then I would go to the Manager, click on the metadata for that file and then do a "usage analysis" on it.

I would enjoy seeing a new Forum on this site into which posters can put down some of their DataStage achievements - not necessarily as personal ones, but showing what results they have achieved with DataStage in their projects recently. I for one love to read about what has been done, how it was ac...

Hello vcannadevula, from your description it looks like you are doing multiple lookups, one per column. This is basic ETL and DataStage technology and is covered in every PDF, sample, tutorial as well as in the forum; so I will assume that this is not what you are asking. Could you perhaps explain a...

Subramaya,

I think it is time to submit a call to Ascential - sounds like an easy issue to reproduce. Please post the response/resolution here as well, since I for one am very interested and I am sure that this scenario will pop up again.

What about 1900-01-01 (with and/or without single quotes)? I just don't like that integer-like look of the date and think that is where the parsing is getting messed up.