large binary data repositories with many files

hi, we have lots of binary data that we are trying to data manage
with subversion. as you can see from the example below svn can be
kind of slow, a commit can take up to 68mins.. are there certain
binary characteristics of files that lend themselves to efficient
diff generation? is there a place i can research the algorithm used
by svn? any ideas how we can speed this up with svn?

               raw
performance repository workarea
           size files(dirs) copy add commit update
size size files(dirs)
           2.1G 5331(2061) 6m 4m 68mins
1.2G 4.2G 44011(22671)

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Wed Apr 12 00:50:33 2006