DSXchange: DataStage and IBM Websphere Data Integration Forum
View next topic
View previous topic
Add To Favorites
Author Message
qt_ky



Group memberships:
Premium Members

Joined: 03 Aug 2011
Posts: 2822
Location: USA
Points: 21387

Post Posted: Mon Feb 09, 2015 11:18 am Reply with quote    Back to top    

DataStage® Release: 11x
Job Type: Parallel
OS: Unix
Additional info: 11.3.1 on AIX
The DataStage job is using the ISD Input and ISD Output stages.

1. ISD is taking the DataStage job columns have a DATE data type and converting them to XML dateTime data type in the WSDL. Example output includes the date, a 'T', the time, and a dash with time zone appended. All we need and want is the date; the data does not contain any time or time zone values.

The only workaround idea I currently have would be to change the job to use Char or Varchar so that ISD would choose an XML string type that contains the date value.

I had expected ISD would have matched the DATE column up a little better by using an XML date data type instead. Is there a way to make ISD force an XML date type?

2. We have noticed that the column order not preserved between the DataStage job and ISD console where you add it as a provider. ISD mixes the column order up randomly. We did not encounter this behavior in ISD 8.7.

Technically, it may not matter once the consumers are automated, but here is an example from an interface documentation and testing standpoint. We have the interface documented with a particular column order, which the job was developed to (i.e. key columns at the top, many output columns following, and status columns and error code column at the bottom). Now a tester or consumer has tediously to hunt through all the output columns searching for the error code, and finds it buried somewhere in the middle.

Is there any way to have ISD preserve the column order?

3. This one is cosmetic in that ISD takes the upper case column names and converts them all to lower case, which we also encountered in 8.7. If you try to override the ISD operation editor, it says, "Variable names must start with a lower case character." It appears that the ISD software cannot be made to comply with corporate naming standards. Is there any way to get around this limitation within ISD?

_________________
Choose a job you love, and you will never have to work a day in your life. - Confucius
eostic

Premium Poster



Group memberships:
Premium Members

Joined: 17 Oct 2005
Posts: 3790

Points: 30485

Post Posted: Mon Feb 09, 2015 4:35 pm Reply with quote    Back to top    

1. Best bet I can think of for number one is to do exactly as you have stated....don't allow ISD to apply any "smarts" to it. Force it to whatever you need, as a character or varchar column, using ...

_________________
Ernie Ostic

blogit!
Open IGC is Here!
Rate this response:  
Not yet rated
qt_ky



Group memberships:
Premium Members

Joined: 03 Aug 2011
Posts: 2822
Location: USA
Points: 21387

Post Posted: Mon Feb 09, 2015 7:02 pm Reply with quote    Back to top    

Thank you for all your responses, Ernie. Very helpful. I may run some of these by Support also.

I should mention that my background is with DataStage and not java or xml. I get xml basics but for the most part am simply relying on ISD to kindly handle xml details for me. So, I would lean towards not taking all that into my own hands, for now! Wink

Regarding question 3, within the xml community, is there a well established naming convention that begins with lowercase? I have noticed a lot of multi-word names in the form of oneTwoThree, whereas our org's traditional convention is in the form of ONE_TWO_THREE. If so, then it would make more sense that ISD is enforcing it.

_________________
Choose a job you love, and you will never have to work a day in your life. - Confucius
Rate this response:  
Not yet rated
ray.wurlod

Premium Poster
Participant

Group memberships:
Premium Members, Inner Circle, Australia Usergroup, Server to Parallel Transition Group

Joined: 23 Oct 2002
Posts: 54407
Location: Sydney, Australia
Points: 295092

Post Posted: Mon Feb 09, 2015 10:21 pm Reply with quote    Back to top    

There is definitely (it seems to me) a trend towards identifiers of the form oneTwoThree, in a number of languages including C, C++, Java and Python. Not so much in databases (yet?).

_________________
RXP Services Ltd
Melbourne | Canberra | Sydney | Hong Kong | Hobart | Brisbane
currently hiring: Canberra, Sydney and Melbourne (especially seeking good business analysts)
Rate this response:  
Not yet rated
eostic

Premium Poster



Group memberships:
Premium Members

Joined: 17 Oct 2005
Posts: 3790

Points: 30485

Post Posted: Tue Feb 10, 2015 4:48 pm Reply with quote    Back to top    

...agreed...and in this case, I suspect it's a java issue and/or something related to the automatic generation of the EAR that is represented by the ISD Application or the plumbing that it uses for ge ...

_________________
Ernie Ostic

blogit!
Open IGC is Here!
Rate this response:  
Not yet rated
Display posts from previous:       

Add To Favorites
View next topic
View previous topic
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum



Powered by phpBB © 2001, 2002 phpBB Group
Theme & Graphics by Daz :: Portal by Smartor
All times are GMT - 6 Hours