> What could be a possible explanation for this behaviour? > One difference int configurations is that the sqlite lib is built for > multithreading in the tmpfs scenario.
Could it be an overhead of mutexes? Make tests with the same SQLite library. Run test with tmpfs using strace to see if filesystem overhead is more than you expect. And post brief results here - it will be interesting to see and/or try to explain them. Pavel On Mon, Aug 8, 2011 at 9:04 AM, <sreekumar...@gmail.com> wrote: > HI, > > I performed an experiment where I do db operations on an in-memory database > and compared it with the results from same operations on a database in > tmpfs.(Same db structure,records etc) > > From the point of view of performance, I expected similar performance , tmpfs > being a little slower due to filesystem overhead. However, the operations on > tmpfs was much slower than expected. > > What could be a possible explanation for this behaviour? > > One difference int configurations is that the sqlite lib is built for > multithreading in the tmpfs scenario. > > -Sreekumar > Sent from BlackBerry® on Airtel > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users