On Wed, 28 Sep 2011 23:40:11 +0800 Mike Buckler <mjbmik...@gmail.com> wrote:
> I have noticed that fossil reads the entire database file (in 1024 > byte increments) several times during the commit process, even when > committing a single file locally with no remote server. > While this doesn't matter for small repos (<10 MB), my test repo at > 40 MB is border line unusable even when run from a fast solid state > disk. The physical directory structure has 5000 files totaling 270MB, > but I don't see any evidence of that being read by the fossil.exe > process, except for the particular file being checked in. > Is this behavior by design? May it be that you're seeing the behaviour described in [1]? Then switching [2] the database to the WAL mode [3] could be a way to go. 1. http://www.sqlite.org/draft/lockingv3.html#rollback 2. http://www.mail-archive.com/fossil-users@lists.fossil-scm.org/msg05742.html 3. http://www.sqlite.org/draft/wal.html _______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
