Hi Everyone, The problem was a mix of processes running SQLite 3.5.3 and 3.3.12. When all processes ran 3.5.3 the problem disappeared and we got a consistent high performance (<4ms per select).
Regards Ken kenmor wrote: > > > Hi Everyone, > > We are seeing the following strange behaviour with SQLite. > > Our environment is as follows > > > 1. SQLite 3.5.3 > 2. Using MEMORY DB > 3. Many databases > 4. Sun Solaris machine T2000 (32 cores and 16G memory) > 5. A single DB table with only 2 entries (so size is not an issue here!) > 6. A production - NOT a debug version of the library. > > > TEST 1 > > A single process. We continually hammer the database with the same query. > We find a row entry in an average of 3ms. > The max response time is approx. 4ms to 5ms. > > TEST 2 > > We run 8 processes (each has its own DB). > The average response time on ALL of the databases increases to 5ms > However the max reponse time has increased to 30ms to 40ms. > > > It appears that some resource is being shared between the processes. > > Can anyone advise us on what resources could be shared and how the problem > could be avoided. > > Any help would be gratefully appreciated > > > > -- View this message in context: http://www.nabble.com/Running-multiple-SQLite-processes-tp14320370p14370260.html Sent from the SQLite mailing list archive at Nabble.com. ----------------------------------------------------------------------------- To unsubscribe, send email to [EMAIL PROTECTED] -----------------------------------------------------------------------------