Electronic MailReading 9.2.1.ppt

Electronic MailReading: 9.2.1 COS 461: Computer Networks Spring 2007 (MW 1:30-2:50 in Friend 004) Jennifer Rexford Teaching Assistant: Ioannis Avramopoulos /courses/archive/spring07/cos461/ Distributed Hash Tables Location-Independent Naming Separate a name from its location File name is “BritneyHitMe.mp3” Current location is Look-up problem Given item’s name, find who has the item Where item stored at dynamic set of nodes Two key design decisions How do we map items on to nodes? How do we route a request to that node? Structured Solution: Special Nodes Central index (Napster) Central server keeps track of who has each file Peers publish a list of their content at the server Peers query the server to locate desired content Simple and efficient, but not scalable or reliable Unstructured Solution: Random Query flooding (Gnutella) Peers do not publish information about content Instead, queries are flooded throughout system And a peer who has the desired content replies Reliable, but long delays and poor scalability Can We Do Better? Decentralized No central coordination Scalable Run with thousands or millions of nodes Fault tolerant Tolerate nodes joining, leaving, failing, … Two key ideas Hashing: to map a name to a value Distributed: for decentralized, self-organization Hashing Name-value pairs (or key-value pairs) E.g., “BritneyHitMe.mp3” and “” E.g,. “Jen Rexford” and “jrex@” Hash table Data structure that associates keys with values Hash Functions Hashing Transform the key into a number And use the number to index an array Example hash function Hash(x) = x mod 101, mapping to 0, 1, …, 100 Challenges What if there are more than 101 nodes? Fewer? Which nodes correspond to each hash value? What if nodes come and go over time? Consistent Hashing Large, sparse identifier space (e.g., 128 bits) Hash a set of keys x uniformly to large id space Hash nodes to the id space as well Where to Store (Key, Value) Pair? Mapping keys in a load-balanced way Store the key at