Need some feed back on a conversion discussion going on here at the Home Office...
I have an open option screen (yea, I know... but, it's just the way it is). The user can pick/chose items that he needs to view. The request pulls data from 2 PF sources. Currently, they are using an OpnQryF to build the selection, then open the file to display.
Tables have about 40k - 50k records
The question is: indexing -OR- SQL based data set returns. If we build LF's across the tables (There are about 4-5 key fields associated to each) then that will push into the CQE and give slower results... No? Then, the question is ... how many to build? Associated access paths included...
My first instinct was to not use indexes (determined on the size of the tables) and build selections in SQL Views or Dynamic SQL as needed BUT (and yes, there's always a big Butt somewhere) what about the linkage between the tables? Should that not be indexed? EVI?
OK, I know most of this is Philosophy... but, just curious on thoughts and reasons why to do one or the other.
Thanks for the feedback!
I have an open option screen (yea, I know... but, it's just the way it is). The user can pick/chose items that he needs to view. The request pulls data from 2 PF sources. Currently, they are using an OpnQryF to build the selection, then open the file to display.
Tables have about 40k - 50k records
The question is: indexing -OR- SQL based data set returns. If we build LF's across the tables (There are about 4-5 key fields associated to each) then that will push into the CQE and give slower results... No? Then, the question is ... how many to build? Associated access paths included...
My first instinct was to not use indexes (determined on the size of the tables) and build selections in SQL Views or Dynamic SQL as needed BUT (and yes, there's always a big Butt somewhere) what about the linkage between the tables? Should that not be indexed? EVI?
OK, I know most of this is Philosophy... but, just curious on thoughts and reasons why to do one or the other.
Thanks for the feedback!
Comment