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 |
|
AskSQLTeam
Ask SQLTeam Question
0 Posts |
 Posted - 2001-03-21 : 18:23:22
|
Michael writes "Hello,
Currently using SQL Server 7.0
We are rewriting our product with SQL Server 2000 so this is what we will be using.
I work on a project where we have a very large amount of text. We actually have 1,250,000 + full text articles which can very from 2-20 pages of text. We also have abstracts and other meta data about these articles which is full text indexed. Currently we have all the text in a field in the db and we have full text indexes on all the fields we are searching on. I am wondering two things. One, do you know of any studies that have tested the scalability of the SQL Servers full text indexing?
Two, can we use a special technique, to increase performance. I would like to create a distributed warehouse environment and put the peek benchmark that the full text engine can handle in each ware house. Then write a searching service that will run separate queries across those ware houses in parallel. However the problem this is obvious, how can you sort all results from each query with relevancy. So, can we setup SQL Server in a special way (i.e. across multi CPUs or Multi Servers) to increase performance. We have a large growth rate and are expecting about 3 million articles by the time the new search engine goes live.
Any advice concerning the design of a proper full text search engine would be greatly appreciated.
Thanx" |
|
|
|
|
|
My Question: Searching - The proper desing for large amounts of full text