[JDEV] A JEP

[Thomas Muldowney]

The flat file XDB doesn't have to be the one you use.  Many people use
xdb_sql or modified file ones for directory hashing.  Take a look here
for both:

http://download.jabber.org/contrib/

--temas


On Tue, 2002-06-04 at 10:55, Ramy Hassan wrote:
> Hi,
> I have a suggestion, I think it is critical, I think the spool files of jabberd need to be created using bigdir algorithm. or may be directory hashes this allow having large number of registered users. I have 80,000 user , yes, few of them are concurrently online , yet my file system can not handle 80,000 file in a single directory.