Mirko, ATM, system metadata consist of three "levels". QuadMapFormats are used to describe transformations of individual SQL values (or tupes of SQL values), QuadMaps refers to QuadMapFormats (via QuadMapValues) and describe some "minimal" independent RDB2RDF transformations, QuadStorages organize QuadMaps. QuadStorages contains only "symlinks" to maps, if you drop a storage you don't drop all mappings inside. If you drop the DefaultQuadStorage or some other built-in thing, it can be safely recovered by DB.DBA.RDF_AUDIT_METADATA ( http://docs.openlinksw.com/virtuoso/fn_rdf_audit_metadata.html ), with first parameter set to 1. This will keep your own data intact. However I recommend to write a script that declares all your formats, RDF Views and storages, to be able to reproduce the configuration after any disasters.
Best Regards, Ivan Mikhailov OpenLink Software http://virtuoso.openlinksw.com On Wed, 2011-02-02 at 14:41 +0100, Mirko wrote: > Hello Ivan, > thanks for clarifying this. A follow up question: is it a good idea to > drop and re-create the DefaultQuadStorage? Will it delete any system > data besides the RDF I stored myself? > > > Thanks, > Mirko > > > 2011/2/1 Ivan Mikhailov <[email protected]> > Hello Mirko, > > All metadata about all RDF storages are kept in "system" graph > <http://www.openlinksw.com/schemas/virtrdf#> (namespace prefix > virtrdf: ) . Subjects of type virtrdf:QuadStorage are RDF > storages. > There are three of them by default: > > SQL> sparql select * from virtrdf: where { ?s a > virtrdf:QuadStorage }; > s > VARCHAR > > _______________________________________________________________________________ > > http://www.openlinksw.com/schemas/virtrdf#DefaultQuadStorage > http://www.openlinksw.com/schemas/virtrdf#DefaultServiceStorage > http://www.openlinksw.com/schemas/virtrdf#SyncToQuads > > 3 Rows. -- 3 msec. > > virtrdf:DefaultQuadStorage is what's in use if no > input:storage > specified. > virtrdf:DefaultServiceStorage will be used for SPARQL > federation. > virtrdf:SyncToQuads is to keep the list of RDF Views that are > translated > into RDB2RDF triggers. > > There are two ways of using your RDF View in SPARQL endpoint > without > define input:storage. > > 1. Create RDF View right in virtrdf:DefaultQuadStorage or add > the view > in other storage and then copy it from there to > virtrdf:DefaultQuadStorage . > In any of these two variants, use > SPARQL ALTER QUAD STORAGE virtrdf:DefaultQuadStorage ... > 2. Use SYS_SPARQL_HOST table as described in > http://docs.openlinksw.com/virtuoso/rdfsparql.html#rdfdefaultgraph > and set SH_DEFINES so it contains your favorite define > input:storage
