Maybe you could benefit of using sqlite3_unlock_notify()? See more info here: http://www.sqlite.org/c3ref/unlock_notify.html.
Pavel On Fri, Aug 7, 2009 at 1:45 PM, Nikolaus Rath<[email protected]> wrote: > Hello, > > I have program that continuously runs several threads (about 4 to 10) > which concurrently read and write to an SQLite database. From > http://www.sqlite.org/sharedcache.html it seems to me that I should be > using shared cache mode. > > Until now, I have set the busy timeout to 500 msec and never had to deal > with SQLITE_BUSY errors. However, there seems to be no corresponding > function for the SQLITE_LOCKED errors that are generated in shared cache > mode. So I changed the code manually sleep for a random amount (0 to 100 > msec) and then try to execute the statement again if it encounters > SQLITE_LOCKED. But now the threads are often waiting for more than 1 > second before they finally get their query executed. > > I suspect this is because the random sleep is wasting a lot of time, > while without shared cache (and with enabled busy timeout) a thread > blocks and gets revived as soon as the lock on the database is > released. > > > How can I avoid this problem? Is there a way to set a busy timeout for > SQLITE_LOCKED as well? Or a more clever method instead of sleeping for > random amounts? Or should I just avoid using shared cache mode? > > > Thanks, > > > -Nikolaus > > -- > »Time flies like an arrow, fruit flies like a Banana.« > > PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6 02CF A9AD B7F8 AE4E 425C > > > _______________________________________________ > sqlite-users mailing list > [email protected] > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
