there used to be an implementation that used a o.a.j.core.fs.FileSystem
but was replaced with the default lucene implementation that directly
writes to and reads from the os filesystem.
see also: http://issues.apache.org/jira/browse/JCR-189
the problem basically was lack of random access support in FileSystem
and definitively with a DB back-end. storing the lucene index in a
database would probably require some sort of replication of the index
files to the local disk to get decent query performance.
regards
marcel
Paco Avila wrote:
Hi, I see that I can store Jackrabbit Filesystem in a database using
DbFileSystem, and the persistence manager using
SimpleDbPersistenceManager. But, can I store the Lucene index in a
database? So, all the files used by Jackrabbit could be stored in a
database.
Thans in advance.
