Distribution–PartIII.pptVIP

Distribution – Part III 7/11 – 2008 Client-Server Distribution with proxies Hierarchical distribution system E.g. proxy caches that consider popularity Popular videos replicated and kept close to clients Unpopular ones close to the root servers Popular videos are replicated more frequently Peer-to-Peer (P2P) Overlay networks P2P Many aspects similar to proxy caches Nodes act as clients and servers Distributed storage Bring content closer to clients Storage limitation of each node Number of copies often related to content popularity Necessary to make replication and de-replication decisions Redirection But No distinguished roles No generic hierarchical relationship At most hierarchy per data item Clients do not know where the content is May need a discovery protocol All clients may act as roots (origin servers) Members of the P2P network come and go P2P Systems Peer-to-peer systems New considerations for distribution systems Considered here Scalability, fairness, load balancing Content location Failure resilience Routing Application layer routing Content routing Request routing Not considered here Copyright Privacy Trading Examples: Napster Napster Program for sharing (music) files over the Internet Approach taken Central index Distributed storage and download All downloads are shared P2P aspects Client nodes act also as file servers Napster Client connects to Napster with login and password Transmits current listing of shared files Napster registers username, maps username to IP address and records song list Napster Client sends song request to Napster server Napster checks song database Returns matched songs with usernames and IP addresses (plus extra stats) Napster User selects a song, download request sent straight to user Machine contacted if available Napster: Assessment Scalability, fairness, load balancing Replication to querying nodes Number of copies increases with popularity Large distributed storage Unavailability of files with low popula