lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Della Bitta <michael.della.bi...@appinions.com>
Subject Re: Many Cores with Solr
Date Tue, 29 May 2012 20:04:06 GMT
In our particular case, we're using this index to do prefix searches
for autocomplete of sparse keyword data, so we don't have much to
worry about on this front, but I do agree that it's a consideration
for those use cases that do reveal information via ranking.

Michael Della Bitta

------------------------------------------------
Appinions, Inc. -- Where Influence Isn’t a Game.
http://www.appinions.com


On Tue, May 29, 2012 at 4:00 PM, Erik Hatcher <erik.hatcher@gmail.com> wrote:
> You do get relevancy related "leakage" though.  With users content all in the same index
and using the same field names, term and document frequencies across the index will be used
for scoring.  This may be (and has been) a good reason to keep separately searchable content
in different indexes/cores.
>
>        Erik
>
>
> On May 29, 2012, at 15:07 , Mike Douglass wrote:
>
>> Thank you.
>>
>> That sounds good - are we sure to get no leakage with this approach?
>>
>> I'd be indexing personal information which must not be delivered without
>> authentication.
>>
>> The solr instance is front-ended by bedework which can handle the auth and
>> adding a query term.
>>
>>> IMO it would be a better (from Solr's perspective) to handle the security
>>> w/ the application code.  Each query could include a "?fq=userID:12345..."
>>> which would limit results to only what that user is allowed to see.
>>
>>
>>
>> --
>> View this message in context: http://lucene.472066.n3.nabble.com/Many-Cores-with-Solr-tp3161889p3986675.html
>> Sent from the Solr - User mailing list archive at Nabble.com.
>

Mime
View raw message