On Thu, Nov 29, 2012 at 12:04 PM, Simon Slavin <[email protected]> wrote:
> I have a folder with 17 independent databases in, each of them opened for > reading an writing occasionally. Two of them have both -wal and -shm > files, even though they shouldn't have been opened for read or write for > days, and the last opening of each one closed connections properly without > crashing. > > Should these journal files exist even when nothing is open or has crashed > ? I thought that properly-closed connections deleted journal files for > databases in WAL mode. > If everybody does sqlite3_close() then those files are deleted on the last call to sqlite3_close(). If they are lingering, that indicates that somebody didn't do their sqlite3_close(). > > Is there some official method to delete them ? I tried VACUUM on one and > when the connection was closed the journal files were deleted, but I don't > want to do that if their existence definitely indicates that my code is not > closing those files correctly. > Simply running: "sqlite3 whatever.db .tables" should be sufficient to delete left-over -shm and -wal files. If that doesn't work, it indicates that some other process still has the database open. > > Simon. > _______________________________________________ > sqlite-users mailing list > [email protected] > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > -- D. Richard Hipp [email protected] _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
