LIKE predicate

[logic4fun]

hi all,
i am running a query which hits a huge table with a billion records and uses a like predicate...here is the query..
select *
from table T
where T.value like '% and %'

i cant ignore the starting % or ending % coz..the value can be anywhere in the string.. like andy...or durand or mandy..so as this particular query is going row by row and scanning the whole database..it is taking hell lot of time..to execute..
IS THERE ANY WAY TO USE SOME ALTERNATIVE FOR "LIKE" PREDICATE..WHICH SPEEDS UP MY QUERY..
thanks in advance..
logic4fun

 

[swampBoogie]

No.

 

[PruSQLer]

SwampBoogie is probably right. I have to ask though - what kind of table is this that has a billion records? Is every record active and thus needs to be scanned or can you eliminate some based on other criteria? Is the column you're querying an indexed one?

 

[fluteplr]

Index is not going to help. Any like with a '%' in the front is going to scan the entire table.

Basically what you have here is a bad database design.