> That's interesting - it would be great if someone could do some test to > compare different CallSignalHandler settings. We could introduce a few > measures like ASR,PDD, jitter and average audio delay. One person said that raising CallSignalHandler number from 1 to 60 gave him 3-5% less CPU usage for 2E1. Me personnaly, I don't think in production I would be ever able to see such a small diviation. > But I can imagine that average dialing phase (from setup to altering - PDD) Are you sure NOW that this is what PDD stands for? ;) > would be longer with CallSignalHandlerNumber=1. Another situation when Why? I can't see any relation. We process Setup and return. How can this influent the time between Setup sent and Alerting received? > single handler will be a bad choice is when accounting/authentication takes > long time (due to unresponsive backend server or some other issues). This is the major issue with auth/accounting modules for now... :( > Anyway, if single threaded design would offer enough performance, there > would be no need to create the gk 2.2... The major goal of 2.2 was a rewrite of single threaded RasServer. Do you remember PendingList and Co. ;))) Proxy part was not affected (and we do discuss this part, not RAS). BTW, it would be nice to have the possibility to measure average audio delay with a couple of other parameters. Volonteers? Andrey. ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ List: Openh323gk-users@lists.sourceforge.net Archive: http://sourceforge.net/mailarchive/forum.php?forum_id=8549 Homepage: http://www.gnugk.org/
