Thank you, Sami. I am so glad to know that I can search without GUI. The command line search returned the result that I expected. So I could be confident tht the index is in good shape.
I did more experiment and find out that I removed too many plugins from search GUI's config file, i.e. $TOMCAT/webapps/ROOT/WEB-INF/classes/nutch-site.xml After restoring the default plugin set, the web GUI is working normally. Thanks a lot. -kuro > -----Original Message----- > From: Sami Siren [mailto:[EMAIL PROTECTED] > Sent: 2006-6-28 9:01 > To: [email protected] > Subject: Re: Zero search result > > Teruhiko Kurosaka wrote: > > >I crawled a very small set of documents (20) off > >my test web site with "nutch crawl" using Nutch nightly > >build from 6/14/06 but replacing hadoop-0.3.3 with > >hadoop-0.2 to work around an obscure bug (NUTCH-266 > >& HADOOP-323). > > > >The crawl didn't show any fatal errors and ended > >peacefully. I verified with Luke that at least > >the index subdirectory has a healthy Lucene index. > >It has terms that I expect. > > > > > Are you trying from commandline or webapp? > > The simplest way to test searching is to run NutchBean from > command line: > > 1. define location of you index (et al) with configuration like > > <property> > <name>searcher.dir</name> > <value>/data/nutch</value> > </property> > > in nutch-site.xml > > 2. enter command > bin/nutch org.apache.nutch.searcher.NutchBean <search term> > > -- > Sami Siren > >
