Home » RDBMS Server » Performance Tuning » Selecivity on like in a query
Selecivity on like in a query [message #133289] Thu, 18 August 2005 07:18
stsy
Messages: 6
Registered: August 2005
Junior Member
Hello
I have a little problem (9204) with a big query where i have a like on a table of 260K rows (where libelle like '%search%'). When i take the explain plan it indicates me a estimated number of rows of 13000 (5% of this table). But when my application user's indicates the search (not a bind variable), they are more accurate : the real number of rows got by this table is about 0.5% and the explain plan is not really good. So is there a solution (by hidden parameters or hint) to indicate in the query the selectivity of like ??

Thanks
Previous Topic: How to make this query fast?
Next Topic: tuning with statpack
Goto Forum:
  


Current Time: Sat Nov 23 16:18:41 CST 2024