lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Karl Wettin <karl.wet...@gmail.com>
Subject Re: Using single or multiple indices for searching different entity types
Date Thu, 08 May 2008 11:47:36 GMT
JP O'Gorman skrev:

> All this information is stored within the one index. All the documents have
> a type and an id field to uniquely identify the entity instance. We allow
> the user to choose what entity types to search over e.g. Just Client and
> Project information or all information etc. We use the type field to specify
> what entity type to restrict the search to. 

Do you use a filter or a term query to restrict the search? Filters 
usually speed things up a lot.

> My question surrounds the best mechanism to index this information? Should I
> be storing these entities in different indices? This would mean I might have
> to preform four searches if searching over all information and maybe lose

MultiReader

> search relevance ordering. The problem with using one index is that I have a
> large amount of fields for each Lucene document that are not populated or
> relevent when the document represents only one of my entities.

FieldSelector


What is your current problem? Do you want to optimize for reponse speed 
or for hardware resources? I think you want a single index. I also think 
you want to benchmark all sort of strategies so you will see what 
affects what.

> Also, I am a Lucene.Net user. Is this still the best place for me to post my
> questions?

We can only help you with basic conceptual things. I don't even know if 
the things I've mentioned exists for Lucene.Net. Look here:

http://incubator.apache.org/lucene.net/


     karl

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org


Mime
View raw message