You can copy it, though you may end up with an index directory that has a mix of old index files and new index files. What you seem to be missing is snapinstaller, which makes Solr see that "aha, look, there is a new index in place, let me open it, warm the searcher I have opened on it, and then throw away the old searcher and expose this new one".
Otis -- Sematext -- http://sematext.com/ -- Lucene - Solr - Nutch ----- Original Message ---- > From: Alex Benjamen <[EMAIL PROTECTED]> > To: solr-user@lucene.apache.org > Sent: Friday, February 29, 2008 3:16:08 PM > Subject: question about snappuller script > > I'm looking at snappuller script and the only thing I see it doing is > managing > the snapshot > pulling via rsync. > > And then once the new distribution is in ${data_dir}/${name}-wip it simpy > moves > it to the > index dir: > > # move into place atomically > mv ${data_dir}/${name}-wip ${data_dir}/${name} > > What if I already have the complete index snapshot on a local disk on the > slave? > Can I simply > move it over to the index directory while solr is running? Is that going to > cause any corruption, > etc... since I don't see anything special that snappuller does after it got > the > new snapshot > > -Alex >
