Thanks for the info. Is the whole result set read into memory meaning that the number of matches I can have have for a query is limited by my machine's memory?
Otis Gospodnetic wrote: > > Hi, > rows=N param just tells Solr how many top N results to return. Solr (and > Lucene, really) still needs to find all documents that match the query and > then score them (and optionally sort them). The more documents and > matches you have, the more time the query will take. > > Otis > > -- > Sematext -- http://sematext.com/ -- Lucene - Solr - Nutch > > ----- Original Message ---- > From: neil22 <[EMAIL PROTECTED]> > To: solr-user@lucene.apache.org > Sent: Monday, April 14, 2008 5:00:05 PM > Subject: solr query time > > > It seems that response time to a query is linear with the size of the > result > set even if I always only want > the first 10 hits back. > > Testing I did - > 1 millions documents that have "feature1" all with the the same score > - query time = 3 seconds to get first 10 hits > 10 millions documents that have "feature1" all with the same score > - query time = 30 seconds to get first 10 hits. > > Is there any optimization I can do so that query time for first 10 is > constant regardless of result set size? > > > -- > View this message in context: > http://www.nabble.com/solr-query-time-tp16678537p16678537.html > Sent from the Solr - User mailing list archive at Nabble.com. > > > > > > -- View this message in context: http://www.nabble.com/solr-query-time-tp16678537p16686732.html Sent from the Solr - User mailing list archive at Nabble.com.
