QTP Concurrent Transaction Model

[Mark Stewart]

Hello, 

Tell me if I'm wrong please.  Your in QTP...here's the scenario:

1...You specify COMMIT AT UPDATE or COMMIT AT n (n = specification number of transactions).  2...You have no OUTPUT statements in your QTP.  
3...You read a relational database file and output to a subfile.

In my opinion this will cause the concurrent model to be invoked thus no exclusive locks will be applied to the relational table because of the database facility possibly going into write mode.  We should only be in read mode if the above is true.  Right?

COMMIT AT RUN however should put us in the consistancy model however but that is another story.

QUESTION...

Will SET FILE <file name> OPEN READ SHARE have any effect here?  I think it would be overkill.  Sorry I don't happen to have a manual in front of me here at home.

I'm looking for others to back me up here or have a different opinion.  At a client of mine, management is hesitant to use QTP to create subfiles but I think that if COMMIT AT UPDATE and there are no OUTPUT statements then there shouldn't be any locking fears.  I know I can use QUIZ, yes but I wanted to clear up any misconceptions of QTP processing.  Thanks.

Mark Stewart
Consultants Club Corp.
Windsor, Ontario
Canada


__________________________________________________________
Get your FREE personalized e-mail at http://www.canada.com