Distributed consensus

PDF Publication Title:

Distributed consensus ( distributed-consensus )

Previous Page View | Next Page View | Return to Search List

Text from PDF Page: 128

128 7.3. EPOCHS BY RECOVERY Algorithm 28: Proposer algorithm with epochs by recovery and fixed quorums. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 state : • Qe: a fixed phase two quorum for each epoch ∀e ∈ E v,emax←nil QP,QA←∅ ife=nilthen e ← 0, v ← γ, goto line 18 else e←e+1 /* Start of Phase 1 for proposal e */ send prepare( e) to acceptors while ∃z∈E:z emax) then emax ← f, v ← w case timeout goto line 1 if v=nilthen v←γ /* Start Phase 2 for proposal (e,v) */ send propose( e,v) to acceptors while QA ̸⊇ Qe do switch do case accept( e ,v) received from acceptor a QA ← QA ∪ {a} case timeout goto line 1 return v In the worst case, the proposals received are equally split between two values associated with the highest epochs. As such we can place the following bound on the cardinality of QP: na −k+1≤|QP|≤2na −2k+1 Table 7.111 shows examples of this relationship between the total number of acceptors 11The number of acceptors for phase one is shown for epoch 1 onwards since the proposer can always

PDF Image | Distributed consensus

PDF Search Title:

Distributed consensus

Original File Name Searched:

UCAM-CL-TR-935.pdf

DIY PDF Search: Google It | Yahoo | Bing

Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info

Cruising Review Topics and Articles More Info

Software based on Filemaker for the travel industry More Info

The Burgenstock Resort: Reviews on CruisingReview website... More Info

Resort Reviews: World Class resorts... More Info

The Riffelalp Resort: Reviews on CruisingReview website... More Info

CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com (Standard Web Page)