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 |
|
sql777
Constraint Violating Yak Guru
314 Posts |
 Posted - 2008-02-12 : 13:00:04
|
| Say you have a table that contains many rows, upto a few million potentially.The table is like:IDTypeID....Basically the table contains different data, differentiated by the TypeID.The data will contain roughly an equal amount of data per 'TypeID', so if there were 1 million rows, and say there are 10 types, each type would have 100K rows.Would it make sense to create 'child phantom' tables that are specific to a particular 'TypeID' ONLY, and do SELECT's from those tables? Basically these tables are used for READS only, any inserts would be done on the main table and then mirrored to the 'child phantom tables'.This is a READ heavy web application, say for every INSERT, there are hundres to thousands of reads. |
|
|
spirit1
Cybernetic Yak Master
11752 Posts |
Posted - 2008-02-12 : 13:01:25
|
| you're talking about horizontal partitioning and this is commonly used._______________________________________________Causing trouble since 1980blog: http://weblogs.sqlteam.com/mladenpSSMS Add-in that does a few things: www.ssmstoolspack.com <- new version out |
 |
|
|
jackv
Master Smack Fu Yak Hacker
2179 Posts |
Posted - 2008-02-18 : 13:17:24
|
| You could further look at splitting the partitions across across different file groups. Although be careful in applying horizontal partitioning , as maintaining your application code could be extra workJack Vamvas--------------------Search IT jobs from multiple sources- http://www.ITjobfeed.com |
|
|
|
|
|
|
Permformance optimization, is this technique done?