On 13 Mar 2019, at 2:31pm, John Smith <[email protected]> wrote: > I am working with IN-MEMORY database. > When my program starts I load data from file-system DB into my IN-MEMORY DB. > All other SQL operations are performed directly on my IN-MEMORY database. > This is in order to keep performance high.
First, make sure you really need to do this. SQLite performance is normally very high, even without taking special measures. You may be wasting programming time and introducing complexity which will be difficult to debug. Run some time-trials. Of course, you may have already run some time-trials. > // Save only intermediate changes (?) > sqlite3_backup_step(p, -1); // Backup all modifications from last time You cannot combine these two things. The Online Backup API backs up an entire database. It does it page by page, without understanding individual rows of data. It cannot select only changes. So you might want to use it, but if you do you'll create a new copy of the entire database every time. You might want to instead use the Resumable Bulk Update extension: <https://sqlite.org/rbu.html> " An RBU Update is a bulk update of a database file that may include many insert, update and delete operations on one or more tables. " Simon. _______________________________________________ sqlite-users mailing list [email protected] http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
