Here is the wiki entry. http://wiki.apache.org/nutch/RunningNutchAndSolr
Using solr was working fine when mapred.job.tracker was set to local when we crawl and run solrindex but when we setmapred.job.tracker to localhost:9001 and use the hadoop task daemons, solrindex gives us many errors saying: org.apache.solr.common.SolrException: Lock obtain timed out: NativeFSLock Presumably, this is because only one process can make a lock and the rest of the child processes just error out. Is there a way to index using multiple processes? Thanks, Steve On Mon, Oct 4, 2010 at 7:22 AM, Thumuluri, Sai < [email protected]> wrote: > Not very clear on the question - but we do use Nutch to crawl and Solr > to index. > > -----Original Message----- > From: Israel [mailto:[email protected]] > Sent: Monday, October 04, 2010 12:02 AM > To: [email protected] > Subject: About SOLR and Nutch > > * Hi, anyone know if integrating SOLR to Nuth, the search can be > improved, > if offered options how search suggestions, etc? > > * I must integrate SOLR, or nutch already does this? >
