1. distributed systems

2. distributed systems and protocols distributed systems: use components located at networked computers use message-passing to coordinate computations distributed system characteristics  concurrency of components  no global clock  components fail independently asynchronous protocols process knows nothing after sending message synchronous protocols paired request and reply (whiteboard diagram)‏

3. remote procedure call (RPC)‏ class of synchronous protocols  process-to-process  may be implemented on or instead of UDP generic RPC functions: name space for remote procedures  flat: needs central authority  hierarchical: per machine, per process,... way to multiplex calls and replies  unique message IDs  boot IDs

4. additional RPC functions: reliable message delivery  ACKs, timeouts  multiple “channels”  concurrency of “open” messages  no “in-order” guarantee fragmentation and reassembly  avoid TCP overhead  no “sliding window”  ACKs & NACKs

5. Lamport's logical clocks operated per process monotonically increasing time counter typically integer, incremented at each event Lamport timestamps event e; process p i ; L i (e); happened-before relation: → rule LC1: increment L i (e) prior to e i rule LC2:  to message m, append time t = L i  [process p j receives message (m,t)]: p j does:  compute L j := max(L j, t)‏  apply LC rule 1  timestamp event e: receive(m)‏

6. global state distributed garbage collection garbage: not referenced within system communication channels, too distributed deadlock detection waits-for relationship distributed termination detection communication channels, too

7. run: total ordering of all events in global history, consistent with each local history ordering → i, for (i=1,2,...,n)‏ linearization (consistent run): a run that is consistent with the → relation on the global history global state predicate: maps domain of global process state to {T,F} stable (garbage, deadlock, termination)‏ unstable (ongoing)‏

8. safety and liveness safety: deadlock is an undesirable predicate P S 0 is system initial state we want P to evaluate to False for all states liveness: termination is a desirable predicate P' for any linearization L, P' evaluates to True, for some state reachable from S 0

9. shared resources critical section (CS) problem: resource shared by distributed processes solution based solely on message passing mutual exclusion solution: requirements ME1 (safety): <= 1 process executing in CS ME2 (liveness): process requests to enter/exit CS eventually succeed

10. mutual exclusion mutex (mutual exclusion object): resource shared by distributed processes two states: {locked, unlocked} arbitration/scheduling: FIFO, priority,... fairness absence of starvation ordering of process entry to CS  who gets the lock next? but, no global clocks... and, recall deadlock problem ME3 (→ ordering): entry to CS is granted in → order

11. peer-to-peer node lookup distributed index of distributed resources example algorithm: Chord –each peer’s IP address hashed to m-bit key hash function is shared among peers text example: 160-bit keys some small fraction of keys appropriated for real nodes –linear ascending search (modulo 2 m )‏ effectively, a circle of 2 m keys –for k, some key successor(k) := key of next actual node of greater key

12. Chord lookup index stored materials by resource name: –hash(name) => key –send (name, file-IP-address) tuple to node(successor(hash(name)))‏ find/retrieve materials by resource name: –hash(name) => key –contact successor(hash(name))‏ ask for (name, file-IP-address) tuple may be multiple records –ask file-IP-address for name materials

13. Chord infrastructure each participating node n –stores IP address of node(successor(key-of-n))‏ –may also store IP address of predecessor –set of resource records submitted by peers {(name, file-IP-address)} –“finger table” for node k m entries: 0,..., (m-1)‏ {(start i, IP-address(successor(start i )))} –start i = (k + 2 i ) mod 2 m

14. Chord lookup example idea: resource advertiser (repository) and resource searcher both hash resource name to one index key search by sending packet around circle without finger tables, n/2 average lookups –pure linear search with finger tables, log 2 n lookups –binary indexing –finger closest predecessor of key(name)‏ look up key 3 at node 1 look up key 14 at node 1 look up key 16 at node 1 search can start anew further around the circle search over: node knows the key sought is between itself and its successor; the node returns successor IP address