Full Text: FASTEST update

Hi,
My users must be able to do a full text search, and add items to the
database if they aren't in yet. But once an item is added, in a lot of cases
they need it directly after it.
But when they search for it after the insert, the article isn't found... It
takes at least 10 seconds before it's found in the full text index...
Is there a way to speed it up? That speed is really very important for my
applciation! I use now "Change tracking" and "Update Index in background".
Should I use incremential population? Without the "update index in
background"? When I don't use "update index in background": does this mean
that it will update the index immediately in the mean-thread (and will be
faster) or that it won't update at all?
Thanks a lot in advance,
Pieter
Pieter wrote on Thu, 9 Feb 2006 12:27:41 +0100:

> Hi,
> My users must be able to do a full text search, and add items to the
> database if they aren't in yet. But once an item is added, in a lot of
> cases they need it directly after it.
> But when they search for it after the insert, the article isn't found...
> It takes at least 10 seconds before it's found in the full text index...
> Is there a way to speed it up? That speed is really very important for my
> applciation! I use now "Change tracking" and "Update Index in background".
> Should I use incremential population? Without the "update index in
> background"? When I don't use "update index in background": does this mean
> that it will update the index immediately in the mean-thread (and will be
> faster) or that it won't update at all?
I don't think you'll get it faster than using Change Tracking with Update
Index In Background. Incremental is used when you want to create all the new
entries since the last full/incremental population, where you don't want
near real-time updates. If you disable Change Tracking then you will have to
schedule Incremental and/or Full populations yourself.
Dan
|||10 seconds is pretty good, especially if your content is binary.
Note that change tracking with update index in background provides near real
time indexing (10s or less). When you don't use this option the changes are
logged in a file and when you issue an update index they are indexed.
Incremental means that each row is retrieved and if it has been
modified/deleted/inserted it is reindexed. This can take as long or longer
than a full population.
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"Pieter" <pietercoucke@.hotmail.com> wrote in message
news:OlyaZvWLGHA.3276@.TK2MSFTNGP09.phx.gbl...
> Hi,
> My users must be able to do a full text search, and add items to the
> database if they aren't in yet. But once an item is added, in a lot of
> cases they need it directly after it.
> But when they search for it after the insert, the article isn't found...
> It takes at least 10 seconds before it's found in the full text index...
> Is there a way to speed it up? That speed is really very important for my
> applciation! I use now "Change tracking" and "Update Index in background".
> Should I use incremential population? Without the "update index in
> background"? When I don't use "update index in background": does this mean
> that it will update the index immediately in the mean-thread (and will be
> faster) or that it won't update at all?
> Thanks a lot in advance,
> Pieter
>
>