SQL Server FTS performance after adding many records -

We have a web application that allows the client to manage large lists of names. For search on these lists, we use FTS of SQL Server 2008, which usually works well. Our largest customer has 9 00,000 names and enjoys sub-other search time.

However, for another new customer, I've recently imported 150,000 names, and the display is terrible (as the server is poorly awful). I checked the full-text indexer and recently claimed to complete a crawl.

Given the execution plans, I notice that in the fast case (for the large customer), SQL Server first frees and then wants the index for the new customer, it wants an index (For new records, 150,000 times, apparently) and then FTS does.

That's why I tried (indicating index (index), but SQL server says that the index does not exist, obviously it does not consider these FTS index as a "real" index. I always use the STM server How can I use FTS?

UPDATE: I tried to reproduce the data, it has no use. Similar performance problems.

Execution Fast performance:

Slow performance:

< P> I tried to add the "Optimized value for unknown" signal at the end of my query (I had to add a dummy variable for the query to reference also in the signal) yet it seems to work well I add the signal I'm a bit upset about, and still looking for a better solution, but now it's working