I am looking to invest in RadControls for ASP.NET Ajax in a month or two but would like clarification of these points please:
1. The database is going to be massive, something close to 20 million rows with 5 columns of data.
The functionality I want to have is this grid style.
In this demo, based on the 300,000 rows of data and applying a filter, it took around 1.6 seconds from filter(search box) to render/display.
Instead of using a filter search box which you have to press enter to apply, I wanted to replace it with this auto-complete search box, one above each column and as the user is typing their search term into this box: the actual grid updates in real-time. The grid narrows again if the user clicks on one of the suggestions brought up by the auto-complete search box (so a filter that also acts as an auto-complete box).
(To clarify the process is: type a word into the search box, the grid updates as if this was a filter box, at the same time a box of suggestions appears matching their search term, if the user adds more letters, the grid responds immediately and updates the filter, the auto-complete box refreshes its suggestions, the grid finally filters down again if the user clicks any of the suggestions).Example: http://www.trirand.net/demoaspnetmvc.aspx
I hope I have made my questions clear enough, I am not yet a customer so I hope I have posted to the right forum, and thank you for helping me clarify these points (I hope you can address each point individually and as clearly as possible, as I really need to be clear myself on these before buying).
Steve
1. The database is going to be massive, something close to 20 million rows with 5 columns of data.
The functionality I want to have is this grid style.
In this demo, based on the 300,000 rows of data and applying a filter, it took around 1.6 seconds from filter(search box) to render/display.
- I could eventually have 30 million rows of data, can you give an estimated time for this size of data to filter and explain the basis of your estimate? It is 100 times the data in the demo, so how many multiples of 1.6 seconds should I expect?
- Can you confirm my understanding is correct that the grid, on first loading, only pulls off the first page of results (and so does not pull of all records at once) but also gets the number of records for use in the pager, and that the filter is applied to the whole data-set and not just the first page of results as this thread seemed to suggest. Could you also clarify why in that thread the Telerik staff said that the whole data-set is pulled to populate the grid for use with filtering? Is this a special case for custom paging only? (I don't really understand the difference).
Instead of using a filter search box which you have to press enter to apply, I wanted to replace it with this auto-complete search box, one above each column and as the user is typing their search term into this box: the actual grid updates in real-time. The grid narrows again if the user clicks on one of the suggestions brought up by the auto-complete search box (so a filter that also acts as an auto-complete box).
(To clarify the process is: type a word into the search box, the grid updates as if this was a filter box, at the same time a box of suggestions appears matching their search term, if the user adds more letters, the grid responds immediately and updates the filter, the auto-complete box refreshes its suggestions, the grid finally filters down again if the user clicks any of the suggestions).Example: http://www.trirand.net/demoaspnetmvc.aspx
- If this dual filter/auto-complete search box arrangement is possible, can the suggestions be limited in the following way to reduce the waiting time: only suggest the closest 10,000 results in order of whole word matches first, for example?
- Based on this setup, what would your time estimate be to display 10,000 suggestions under these conditions given that the data to search is 30 million rows deep?
I hope I have made my questions clear enough, I am not yet a customer so I hope I have posted to the right forum, and thank you for helping me clarify these points (I hope you can address each point individually and as clearly as possible, as I really need to be clear myself on these before buying).
Steve