On Mon, Apr 27, 2009 at 10:40 PM, hpn1975 nasc <hpn1...@gmail.com> wrote:
> > 1- Guarantee that my searcher (solr) ALWAYS search in my index in *memory > * (use RAMDirectory). Not to use cache. It is possible to disable all caches. But it is not possible to use RAMDirectory right now. This is in progress. https://issues.apache.org/jira/browse/SOLR-465 > > 2- Guarantee that my searcher (solr) ALWAYS search in my index in *file > system* (use FSDirectory). Yes, that is the default and only way currently. > > 3- Persist my index is genereted in only one archive in File System > (optimized) The useCompoundFile setting in solrconfig.xml can help here. > > 4- Persist my index (RAMDirectory) in serializable archive java. I need > create a loader that load my .ser and deseriaze the class RAMDirectory e > set > in the searcher class. I don't think you can use RAMDirectory right now. However if the use-case behind serializing a ram directory is for replication, then there are alternate methods available. http://wiki.apache.org/solr/CollectionDistribution http://wiki.apache.org/solr/SolrReplication > Is possible add any component that manipule the index > ? > Yes. You can write your own request handlers and search components. -- Regards, Shalin Shekhar Mangar.
