Hi, Sounds like you don't need to reindex. You need to find duplicates and delete them.
Otis -- SOLR Performance Monitoring - http://sematext.com/spm On Dec 11, 2012 12:42 PM, "蒋明原" <mailtojiangmingy...@gmail.com> wrote: > Thank you,first of all, > Yes,no same unique key means no this trouble. > But for me now,I can't reindex my data,it's too big.And it,s in production > environment . > So,any friends have solutions? > > Thank you . > > On Wednesday, December 12, 2012, Pawel wrote: > > > I think that solution is quite obvous. Be sure that you don't have items > > with the same unique key in many shards :) > > > > On Tue, Dec 11, 2012 at 5:24 PM, 蒋明原 <mailtojiangmingy...@gmail.com > <javascript:;>> > > wrote: > > > > > hi,all, > > > > > > I'm doing a distribute facet query,and there duplicate data among the > > > distribute cluster. > > > for example: > > > > > > server A hold documents: > > > > > > Doc1: uniqueKey=1 userid=a > > > Doc2: uniqueKey=2 userid=b > > > Doc3: uniqueKey=3 userid=c > > > > > > server B hold documents: > > > Doc1: uniqueKey=1 userid=a > > > Doc2: uniqueKey=4 userid=b > > > Doc3: uniqueKey=5 userid=c > > > > > > when a make a facet query using filed "userid", the expect result is: > > > a:1 > > > b:2 > > > c:2 > > > > > > but solr gives me: > > > > > > a:2 > > > b:2 > > > c:2 > > > > > > However, I make a normal query using : userid:a, > > > solr gives me total 1 result. > > > > > > It seems like: when making facet query,duplicate key will still > > participate > > > in calculate,but when making normal query,solr will choose only 1 > > document > > > between duplication document. > > > > > > So,My problem is "how to remove duplicate documents during distributed > > > facet search." > > > > > > thanks ! > > > > > >
