You just address your specific node and append "&distrib=false" to the URL (or set the param in SolrJ). Then the request goes to the node you've addressed and doesn't get distributed to the other shards.
Best Erick On Fri, Feb 15, 2013 at 10:15 AM, o.mares <ota.mares+nab...@gmail.com>wrote: > Mark Miller-3 wrote > > On Feb 15, 2013, at 6:04 AM, o.mares < > > > ota.mares+nabble@ > > > > wrote: > > > >> Hey when running a solr cloud setup with 4 servers, managing 3 cores > each > >> splitted on 2 shards, what are the proper steps to do a full index > >> import? > >> > >> Do you have to import the index on all of the solr instances? Or is it > >> sufficient enough to perform the import on one instance and it will get > >> replicated automagically? What about core reloading after performing an > >> import, is it necessary to reload the cores? > > > > Yes you can send to one instance, no, you don't need to reload. > > > >> > >> And is there a way to verify that the other instances got the new > indexed > >> data? > > > > Query each node with distrib=false. > > > >> > >> Ota > >> > >> > >> > >> -- > >> View this message in context: > >> > http://lucene.472066.n3.nabble.com/solr-4-1-cloud-and-dataimport-tp4040602.html > >> Sent from the Solr - User mailing list archive at Nabble.com. > > Could you please be more specific about "Query each node with > distrib=false. > "? > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/solr-4-1-cloud-and-dataimport-tp4040602p4040635.html > Sent from the Solr - User mailing list archive at Nabble.com. >
