See below... On Oct 17, 2011, at 11:15 AM, lorenlai wrote:
> 1) I would like to know if it is possible to import data (feeding) while > Solr is still running ? Yes. You can search and index new content at the same time. But typically in production systems you may have one or more "master" SOLR instances accepting new documents, and then setup SOLR replication and multiple "slave" instances behind a load balancer in order to handle search requests. > 2) How can I import my data as into the index ? Via HTTP ? And is it > possible to "automate" this feeding process ? You can post new documents via HTTP POST (as single documents, or as a batch of documents). You can also use a data import handler (DIH) to pull data from some repository such as a SQL database, and then initiate such imports via HTTP request called by a cron job, etc. > > Any LINKS regarding to this topic? :-) http://wiki.apache.org/solr/UpdateXmlMessages http://wiki.apache.org/solr/DataImportHandler > > 3) Is it possible to write a "Batch-Loader" (Batch Jobs) which import the > data into the index ? Yes. You can use any of the available SOLR clients from some program or script (solrj for java, various python clients, etc.). Or you could write some java code that embeds SOLR (or even Lucene) directly to build indexes. There are many possibilities in that case. > > Any LINKS regarding to this topic? :-) http://wiki.apache.org/solr/IntegratingSolr
