Jae, Those numbers are not huge for modern type hardware. Are you actually seeing performance problems? That said, if you are simply spreading your 1 index on 1 server over N cores on the same 1 server, and that's what it sounds like, I am not sure if you are really benefiting.
Otis -- Sematext -- http://sematext.com/ -- Lucene - Solr - Nutch ----- Original Message ---- > From: Jae Joo <[EMAIL PROTECTED]> > To: solr-user@lucene.apache.org > Sent: Friday, May 16, 2008 4:05:29 PM > Subject: Re: Multi core weight > > Running multiple indivisual queries is one option, but because of the volume > of documents (14 millions) and traffic, 10 request per second, I am looking > the optimal way to do that. > > Thanks, > > Jae > > On Thu, May 15, 2008 at 11:57 AM, Otis Gospodnetic < > [EMAIL PROTECTED]> wrote: > > > Jae, > > It sounds like you are doing a distributed search across your 3 cores on a > > single Solr instance? Why not do run 3 individual queries (parallel or > > serial, your choice) and pick however many hits you need from each result? > > > > Otis > > -- > > Sematext -- http://sematext.com/ -- Lucene - Solr - Nutch > > > > > > ----- Original Message ---- > > > From: Jae Joo > > > To: solr-user@lucene.apache.org > > > Sent: Thursday, May 15, 2008 9:07:54 AM > > > Subject: Multi core weight > > > > > > Hi, > > > > > > I am looking for "the best or possible" way to add WEIGHT to each core in > > > multi core environment. > > > > > > core 1 has about 10 millions articles from same publisher and core 2 and > > 3 > > > have less than 10k. > > > I would like to have BALANCED Query result - ex. 10 from core 1, 10 from > > > core 2 and 10 from core 3.. > > > > > > Thanks, > > > > > > Jae > > > >
