"Gene Allen" <[EMAIL PROTECTED]> wrote in message news:!&[EMAIL PROTECTED] > I'm using SQLite v. 3.5 in my multithreaded application. > > I have 2 threads both writing to the same database, different tables > but same database. All is good until one of the threads goes under > HEAVY load, then the other thread blocks never to return (on a bind > statement).
The two threads are using the same connection, right? Each connection has a mutex that SQLite locks on every API call involving this connection. > Is opening the database in each thread, really the only thing I need > to worry about when I have mutilple threads are using the same > database file, as long as that are using seperate tables? No. Multiple connections (whether from different processes or within the same process) essentially maintain a multiple-readers-single-writer lock on the whole database, not per table. However, you can achieve a sort of table-level locking by enabling shared cache: http://sqlite.org/sharedcache.html Igor Tandetnik _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
