On 3/27/07, Luigi Lauro <[EMAIL PROTECTED]> wrote:
On 27 Mar 2007, at 14:51, Oystein Grovlen - Sun Norway wrote: > An empty database with just system tables etc has less than 70 > files/directories. > Then you will use a few log files and maybe some property files so > I would say that you will have about 175 files available for your > own data. Is it possible to cut down theses numbers both by # of files and filesize? By disabling all the unneeded logging, disabling features, forcing to write all logs to the same file, whatever. ANYTHING is fine, if it cuts down number of files and filesize. > The database will use one extra file for each table and index you > create. Ok, so file number grow linearly with the # of tables in the database. Is there a way to reduce this number, by storing several/ all tables into one 'data' file? > Do you need to do backups to the same system? In that case, you > will need to be able to store more than one copy of each file. No, I don't need backup. This is a 'cache' of a remote database, so I don't need ANYTHING at all, I just need to keep file number and file size to the MINIMUM possible with Derby :)
Then I wonder if this feature (in-memory storage) would help solving your problem - at least, in respect with the number of generated files / directories: https://issues.apache.org/jira/browse/DERBY-646 Thanks a lot for the help :)
