Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
dkperez
Starting Member
5 Posts |
 Posted - 2008-08-20 : 09:05:42
|
| Traffic question... I'm being told by the folks that built the web page that the front-end is only getting the records it needs, and I'm curious how this is happening...Situation is that several hundred records are retrieved from the database, but they are to be displayed 20 per screen on the front end. (IGNORE the idea of retrieving them 20 at a time from the database...)If I retrieve 400 records from a database table in a subroutine called from the front-end code, do the records get pushed across the network? Does the front-end somehow communicate with the database to only pull or accept a subset of the records? If the user selected page 19 to view, would the front-end somehow only grab records 361-380 or would at least 380 records get sent to the front end for processing? I'm being told by the folks that built the web page that the front-end is only getting the records it needs, and I'm curious how this is happening... |
|
|
spirit1
Cybernetic Yak Master
11752 Posts |
Posted - 2008-08-20 : 10:12:15
|
| doing paging on the client means that on the first call you get all 400 rows of data, the client caches them and does paging on the cached data.so no, there's no "smart" communication between the client and the server for this._______________________________________________Causing trouble since 1980Blog: http://weblogs.sqlteam.com/mladenpSpeed up SSMS development: www.ssmstoolspack.com <- version 1.0 out! |
 |
|
|
dkperez
Starting Member
5 Posts |
Posted - 2008-08-20 : 11:37:46
|
| Thanks for the quick reply..... I've rummaged around and found out they're using an ASP ADO recordset. Then to get the page of data theydo a "move" to the starting point and read from there.....SO, the user's browser isn't getting hit with all the records, but the whole recordset is getting dragged outof the database and shoved over to where the interface software is runnig, correct?SO, the question is, is it SIGNIFICANTLY faster and more tolerant of multiple users to have the databaseretrieve only what's needed or to do this as it's currently being done? I'm not much concerned when there'sjust one user working, but what happens when 1000 users are pulling 300 or 500 thousands of records? |
|
|
|
|
|
|
|
|
How many records from DB to frontend?