On Mon, Dec 13, 2010 at 12:29:20PM +0200, Yoni Londner wrote: > Hi, > > In general I think that SQLite should have a in memory VFS, which is OS > independent.
A laudable goal, where mmap is not available. > > I am going to implement proc_exclusive now, and would love to get any > directions/tips from sqlite developers (if you think its not a good > idea, please let me know, or better - suggest another way :-)) While I'm not a SQLite developer, I have to say I think you're going down the wrong path. Are you sure the WAL index is your bottleneck? I think it is unlikely. >From your previous descriptions of your application, you might be better off rethinking how big you make your batches, and committing more often than you currently do, and using the default checkpoint mechanism in a single thread. With smaller batches, checkpointing will take less time, so you'll see less jitter when inserting data. Or, have smaller batches with a background checkpointing thread. As it is at the moment, your batches sound like they're too big, and the checkpoint thread can never hope to catch up, hence your huge checkpoint file. In order to help further, you'd need to specify more details of the event source, such as how big and how often data is coming in. Christian _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
