Broadcast Algorithms BJRN A. JOHNSSON Overview Best-Effort - - PowerPoint PPT Presentation

▶
broadcast algorithms
SMART_READER_LITE
LIVE PREVIEW

Broadcast Algorithms BJRN A. JOHNSSON Overview Best-Effort - - PowerPoint PPT Presentation

Dec 02, 2023 48 likes •440 views

Broadcast Algorithms BJRN A. JOHNSSON Overview Best-Effort Broadcast (Regular) Reliable Broadcast (2) Uniform Reliable Broadcast (2) Stubborn Broadcast Logged Best-Effort Broadcast Logged Uniform Reliable Broadcast

slide-1
SLIDE 1

Broadcast Algorithms

BJÖRN A. JOHNSSON

slide-2
SLIDE 2

Overview

  • Best-Effort Broadcast
  • (Regular) Reliable Broadcast (2)
  • Uniform Reliable Broadcast (2)
  • Stubborn Broadcast
  • Logged Best-Effort Broadcast
  • Logged Uniform Reliable Broadcast
  • Probabilistic Broadcast (2)
slide-3
SLIDE 3

Abstracting Processes

  • Crash-stop
  • Omissions
  • Crash-recovery
  • Eavesdropping faults
  • Arbitrary-fault (Byzantine)
  • Faulty or correct
slide-4
SLIDE 4

Timing Assumptions

  • Asynchronous System
  • Synchronous System
  • Partial Synchrony
slide-5
SLIDE 5

Distributed-System Models

  • Fail-stop

– crash-stop, perfekt links, perfect failure detector (P)

  • Fail-noisy

– crash-stop, perfekt links, eventually P (P)

  • Fail-silent

– crash-stop, perfekt links, no failure detector

  • Fail-recovery

– crash-recovery, stubborn links, eventual leader detector (Ω)

  • Fail-arbitrary

– fail-arbitrary, authenticated perfekt links

slide-6
SLIDE 6

Motivation

  • Client-server scheme – point-to-point communication

– Useful when reliable, e.g. TCP

  • Bigger systems usually more than 2 processes

– Broadcast abstractions convenient – Send to all processes, in a single one-shot op.

  • Reliability req. of p2p not directly transposable

– “No message lost or duplicated” – Complex for broadcast…

slide-7
SLIDE 7

Best-Effort Broadcast

  • Burden of ensuring reliability on sender:

– Receivers unconcerned with enforcing reliability – No delivery guarantees if sender fails

slide-8
SLIDE 8

Best-Effort Broadcast

PP2PL

slide-9
SLIDE 9

Best-Effort Broadcast

slide-10
SLIDE 10

Best-Effort Broadcast

slide-11
SLIDE 11

(Regular) Reliable Broadcast

  • Best-effort only ensures delivery if sender doesn’t crash

– Processes might not agree on message delivery – Even if all messages sent before sender crashes…

  • (Regular) Reliable broadcast provides stronger notion of

reliablity: – Ensures agreement even if sender fails. – Sender failure – no process delivers message

slide-12
SLIDE 12

(Regular) Reliable Broadcast

New!

slide-13
SLIDE 13

(Regular) Reliable Broadcast

array of sets

  • riginal source

message descriptor

slide-14
SLIDE 14

(Regular) Reliable Broadcast

slide-15
SLIDE 15

(Regular) Reliable Broadcast

slide-16
SLIDE 16

(Regular) Reliable Broadcast

slide-17
SLIDE 17

(Regular) Reliable Broadcast

  • Problem: only requires the correct processes deliver the

same set of messages

slide-18
SLIDE 18

Uniform Reliable Broadcast

Different!

slide-19
SLIDE 19

Uniform Reliable Broadcast

Infinite array for all possible message…

slide-20
SLIDE 20

Uniform Reliable Broadcast

slide-21
SLIDE 21

Uniform Reliable Broadcast

slide-22
SLIDE 22

Uniform Reliable Broadcast

  • Requires N > 2f

Size of ack[m]

slide-23
SLIDE 23

Stubborn Broadcast

No duplication gone!

slide-24
SLIDE 24

Stubborn Broadcast

slide-25
SLIDE 25

Logged Best-Effort Broadcast

  • First for Fail-recovery model
  • Strongest model, uniform reliable, not enough
  • Difficulty: crashing, recovery and never crashing again is

correct

  • Solution: stable storage, as seen in “logged perfect links"
slide-26
SLIDE 26

Logged Best-Effort Broadcast

slide-27
SLIDE 27

Logged Best-Effort Broadcast

ensures Validity

slide-28
SLIDE 28

Logged Uniform Reliable Broadcast

New!

slide-29
SLIDE 29

Logged Uniform Reliable Broadcast

slide-30
SLIDE 30

Logged Uniform Reliable Broadcast

slide-31
SLIDE 31

Probabilistic Broadcast

  • No deterministic broadcast guarantees
  • Offers “cost” reduction at the price of lower reliability
  • a. Reliability not scalable – ack implosion problem
  • b. Possible solution – requires configuration
  • Epidemic dissemination – rumor spreading, gossiping
slide-32
SLIDE 32

Probabilistic Broadcast

Weaker than validity

slide-33
SLIDE 33

Eager Probabilistic Broadcast

  • Sends to k random processes – the fanout
  • A round of gossiping = receiving and resending message
  • R rounds of gossiping per message
  • R and k determine efficiency of algorithm
slide-34
SLIDE 34

Eager Probabilistic Broadcast

Returns k processes from Π \ {self}

slide-35
SLIDE 35

Lazy Probabilistic Broadcast

  • EPB too eager; consumes resources and causes

redundant transmissions

  • 1. Gossip until e.g. N/2 processes infected (push-phase)
  • 2. Missed processes ask for message (pull-phase)
slide-36
SLIDE 36

Lazy Probabilistic Broadcast

slide-37
SLIDE 37

Lazy Probabilistic Broadcast

slide-38
SLIDE 38

Lazy Probabilistic Broadcast

slide-39
SLIDE 39