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 |
|
dextrous
Starting Member
15 Posts |
 Posted - 2010-01-09 : 00:58:44
|
| Hello,This question is in relation to zipcodes and indexing. I have a database of zip+4 connected to cities (about 24 million rows). I have a cluster index over the zip+4.When my program retrieves data, it is much faster when it retrieves the earlier numbers (which even to a newbie like me makes sense since it comes up first) and takes 2-3 times longer if the zip is say, 99999-9999.I created views (5) on the index to divide it up but the problem seems to persist.Any ideas on what might be the most efficient way to index zipcodes? |
|
|
russell
Pyro-ma-ni-yak
5072 Posts |
Posted - 2010-01-09 : 01:12:43
|
| show the DDL and the Queries Also, set statistics io on and show us the results of that too. |
 |
|
|
jackv
Master Smack Fu Yak Hacker
2179 Posts |
Posted - 2010-01-09 : 06:53:34
|
| As well as information requested, consider usinng a non-clustered index - particuarly if selectivity is high - which would mean the data is relatively unique. But this is dependant on whatr type of query you are making. For example , if the zip codes are linked to cities - would I be right in thinking these are quite random seaeches?Jack Vamvas--------------------http://www.ITjobfeed.com |
|
|
|
|
|
|
Which index to use?