Hi In the blog post http://heartbeat.skype.com/2007/08/the_microsoft_connection_explained.html It said "This combination of factors created a situation where the self-healing needed outside intervention and assistance by our engineers."
I'm curious about how they can intervene and assist the core netwrok self-healing. As far as I konw, Skype select supernode automatically, and user cann't control whether his/her skype client become a supernode. They means they put MASS their dedicated machines into the core network, make them be supernodes, or they can ONLY modify some parameters through network communiction to control how supernodes are selected, without update client program ? 2007/8/22, Alen Peacock <[EMAIL PROTECTED]>: > An update: > http://heartbeat.skype.com/2007/08/the_microsoft_connection_explained.html > > Still too vague to really learn anything specific about how to prevent > something like this from happening anywhere outside of Skype, but it > does try to address #2 of zooko's questions (and very vaguely #3). > > If I'm not misinterpreting, it does sound now like this was an issue > of too many supernodes going away at once. Is that what he is > referring to when he says "Skype's peer-to-peer core?" From later use > of "a combination of high load and supernode rebooting" it seems to > be. I'm curious if Skype's supernode population is more heavily skewed > towards Windows boxes than the general Skype population? Does Skype > not have very many mac/linux users, or are mac/linux users less likely > to be chosen as supernodes? And what in the world is referred to by > "P2P network resource allocation algorithm Skype used?" > > Alen > _______________________________________________ > p2p-hackers mailing list > [email protected] > http://lists.zooko.com/mailman/listinfo/p2p-hackers > _______________________________________________ p2p-hackers mailing list [email protected] http://lists.zooko.com/mailman/listinfo/p2p-hackers
