Crashed router. Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concepts and Design Edn. 3 Pearson Education

Transcription

1 Figure 11.1 A network artition Crashed router Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

2 Figure 11.2 Server managing a mutual exclusion token for a set of rocesses Queue of requests Server Grant token 1 1. Request token 2. Release token Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

3 Figure 11.3 A ring of rocesses transferring a mutual exclusion token 1 2 n 3 4 Token Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

4 Figure 11.4 Ricart and Agrawala s algorithm On initialization state := RELEASED; To enter the section state := WANTED; Multicast request to all rocesses; Request rocessing deferred here T := request s timestam; Wait until (number of relies received = (N 1)); state := HELD; On receit of a request <T i, i > at j (i j) if (state = HELD or (state = WANTED and (T, j ) < (T i, i ))) then queue request from i without relying; else rely immediately to i ; end if To exit the critical section state := RELEASED; rely to any queued requests; Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

5 Figure 11.5 Multicast synchronization Rely Rely 34 Rely 2 34 Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

6 Figure 11.6 Maekawa s algorithm (cont d on next slide) On initialization state := RELEASED; voted := FALSE; i For to enter the critical section state := WANTED; Multicast request to all rocesses in V i { i }; Wait until (number of relies received = (K 1)); state := HELD; On receit of a request from i at j (i j) if (state = HELD or voted = TRUE) then queue request from i without relying; else send rely to i ; voted := TRUE; end if Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

7 Maekawa s algorithm (cont d) i For to exit the critical section state := RELEASED; Multicast release to all rocesses in { }; V i On receit of a release from i at j (i j) if (queue of requests is non-emty) then remove head of queue from send rely to k ; voted := TRUE; else voted := FALSE; end if i k, say; Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

8 Figure 11.7 A ring-based election in rogress Note: The election was started by rocess 17. The highest rocess identifier encountered so far is 24. Particiant rocesses are shown darkened Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

9 Figure 11.8 The bully algorithm election Stage 1 C election answer answer election Stage 2 election answer election C timeout Stage Eventually... Stage 4 coordinator C The election of coordinator 2, after the failure of 4 and then 3 Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

10 Figure 11.9 Oen and closed grous Closed grou Oen grou Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

11 Figure Reliable multicast algorithm On initialization Received := {}; For rocess to R-multicast message m to grou g B-multicast(g, m); // g is included as a destination On B-deliver(m) at rocess q with g = grou(m) if ( m Received ) then Received := Received { m} ; if ( q ) then B-multicast(g, m); end if R-deliver m; end if Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

12 Figure The hold-back queue for arriving multicast messages Message rocessing deliver Hold-back queue Delivery queue Incoming messages When delivery guarantees are met Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

13 Figure Total, FIFO and causal ordering of multicast messages T 1 T 2 F 1 F 2 F 3 Time C 1 C 2 C 3 P 1 P 2 P 3 Notice the consistent ordering of totally ordered messages T 1 and T 2, the FIFO-related messages F 1 and F 2 and the causally related messages C 1 and C 3 and the otherwise arbitrary delivery ordering of messages Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

14 Figure Dislay from bulletin board rogram Bulletin board: os.interesting Item From Subject 23 A.Hanlon Mach 24 G.Joseh Microkernels 25 A.Hanlon Re: Microkernels 26 T.L Heureux RPC erformance 27 M.Walker Re: Mach end Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

15 Figure Total ordering using a sequencer 1. Algorithm for grou member r g On initialization: := 0; To TO-multicast message m to grou g B-multicast( g { sequencer( g) }, <m, i>); On B-deliver(<m, i>) with g = grou(m) Place <m, i> in hold-back queue; On B-deliver(m order = < order, i, S>) with g = grou(m order ) wait until <m, i> in hold-back queue and S = r g ; TO-deliver m; // (after deleting it from the hold-back queue) r g = S + 1 ; 2. Algorithm for sequencer of g On initialization: := 0; s g On B-deliver(<m, i>) with g = grou(m) B-multicast(g, < order, i, s g >); s g := s g + 1; Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

16 Figure The ISIS algorithm for total ordering RM1 P Message 2 Proosed Seq 3 Agreed Seq RM2 P P 1 FE P 3 RM3 Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

17 Figure Causal ordering using vector timestams Algorithm for grou member i ( i = 12,, N ) On initialization g V i [ j] := 0 ( j = 12,, N ); To CO-multicast message m to grou g g g V i [] i := V i [] i + 1 ; g B-multicast(g, < V i, m>); On B-deliver(<, m>) from, with g = grou(m) g j lace < V j, m> in hold-back queue; g g g g wait until V j [ j] = V i [ j] + 1 and V j [ k] V i [ k] ( k j ); CO-deliver m; // after removing it from the hold-back queue g g V i [ j] := V i [ j] + 1 ; V j g Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

18 Figure Consensus for three rocesses P 1 d 1 :=roceed d 2 :=roceed P 2 v 1 =roceed 1 v 2 =roceed RM2 Consensus algorithm v 3 =abort RM3 P 3 (crashes) Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

19 Figure Consensus in a synchronous system i Algorithm for rocess g; algorithm roceeds in f + 1 rounds On initialization 1 0 Values i := { }; Values i = {}; v i In round r ( 1 r f + 1) r r 1 B-multicast(g, Values i Values ); // Send only values that have not been sent r + 1 r i Values i := Values i ; while (in round r) { On B-deliver( V j ) from some r + 1 r + 1 j Values i := Values i V j ; } After ( f + 1) rounds f + 1 Assign d i = minimum( Values i ); Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

20 Figure Three byzantine generals 1 (Commander) 1 (Commander) 1:v 1:v 2:1:v 2 3 3:1:u 1:w 1:x 2:1:w 2 3 3:1:x Faulty rocesses are shown shaded Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education

21 Figure Four byzantine generals 1 (Commander) 1 (Commander) 1:v 1:v 1:v 2:1:v 2 3:1:u 3 4:1:v 4:1:v 2:1:v 3:1:w 1:u 1:w 1:v 2:1:u 2 3:1:w 3 4:1:v 4:1:v 2:1:u 3:1:w 4 Faulty rocesses are shown shaded 4 Instructor s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concets and Design Edn. 3 Pearson Education