From your documentation and also some posts here I noticed that "binding to IQueryable" can be use enhance performance.
Here for an example http://www.telerik.com/forums/enablecaching-=-true-causing-pivotgrid-to-remain-empty#lN8o0P-F3UKixSLep-_wbw you link
to this documentation entry
http://docs.telerik.com/devtools/aspnet-ajax/controls/pivotgrid/data-binding/advanced-via-needdatasource
What I see there is that it all ends up with a "select everything from a table" - "SELECT * FROM Transportation"
What I EXPECT is a select statement which "fits my needs".
I know sometimes there is a need to select all data.
But (for an example) if I filter for Transportation Items which happend in the year 2010 and have a value bigger than 10.000 I would expect a query like
"SELECT * FROM Transportation WHERE TPYear = 2010 AND TPValue>10.000"
Did I miss something?
By the way - I also tried to use an "EntityDatasource" (using EF6) - which does the same - SELECT * FROM....
No filtering on the query occures.
On last thing - in the first linked post the question contains: .... We are running on a decently large initial result set (about 3000 records) ....
This is not (in our case) what I'd call "large" - consider the query product of (only) 1000 sales per day with articels, regions,.. for (only) 3 years.
Which also makes the star in the select questionable - why select * if I only want to know ID and quantity (for an example).
Am I thinking totaly wrong - or is there something I've overseen?
Manfred
1 Answer, 1 is accepted
The RadPivotGrid requires the entire data set in order to perform its filtering and any other data operation on that matter (like paging, sorting, etc.). In regards of performance, usingEntityDataSource should be much faster than the advanced data-binding, but the best performance should be with OLAP cube:
Note that in the same manner, the RadGrid requires the entire data set in order to perform its data operations.
Hope this helps.
Regards,
Konstantin Dikov
Telerik by Progress