lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From John Patterson <>
Subject Re: Index types
Date Wed, 27 Aug 2008 11:35:56 GMT

It is a blurry line between the need to use a DBMS and lucene.  For now
Lucene works fine for my search needs and I use Terracotta to persist
application state.  So no need for a DBMS at all currently - although under
the hood Terracotta uses BDB JE.

Does Solr's range impementation use the large Boolean SHOULD queries?

Karsten F. wrote:
> Hi John,
> about "integration other index implementation":
> Sounds like you need a DBMS with some lucene features.
> There was a post about using lucene in Oracle:
> and
> But normaly for a particular problem there is a solution with lucene
> (example: you should not index timestamps in lucene, but you can A) index
> year, month and day  or   B) generate your own filter which used the
> (cached) timestamp from a stored field  or  C) use solr (which contains
> timestamp-range implementation out of the box).
> Best regards
>    Karsten
> John Patterson wrote:
>> Hi, I know that Lucene uses an inverted index which makes range queries
>> and great-than/less-than type queries very slow for continuous data types
>> like times, latitude, etc.  Last time I looked they were converted into
>> huge OR queries and so had a maximum clause limit.  
>> I was wondering if any work had been done on integrating other index
>> implementations?
>> Thanks,
>> John

View this message in context:
Sent from the Lucene - Java Users mailing list archive at

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message