137 questions
-3
votes
1
answer
52
views
Why does Cassandra need consensus when we can use read-repairs?
As per DDIA (Design data intensive application) to make writes atomic in distributed databases we can use read repair which means lets say if client made read request to 3/5 replica nodes and on two ...
0
votes
1
answer
138
views
Doesn't Paxos end up with the same instructions in the exact same order?
I am trying to understand why ZAB is required when Paxos does the same thing. I read multiple documentations and articles and all summed up that Zookeeper requires the instructions to be in the exact ...
2
votes
2
answers
307
views
Paxos algorithm: Dependency of Accept and Prepare phases
Paxos algorithm generally represented in 2 phases:
Prepare phase
Propose (or Accept) phase
Each phases request are send to Acceptors which making decisions. All requests are equipped with unique ids ...
1
vote
0
answers
458
views
Pax device NFC support
As I am working on a pax project which is android based device. Devices are Pax A920 pro and Pax A920. Pax provide SDKs for these devices and for NFC reader of there is only support provided for ...
2
votes
1
answer
351
views
Cassandra LWT CasWriteUnknownResultException failure status is unclear
When executing a LWT such as INSERT .... IF NOT EXISTS, the Java driver may throw a CASWriteUnknownException. It's unclear (a) what phase failed (b) if the LWT transaction has completely failed or was ...
0
votes
1
answer
135
views
How Raft know previous term log entry committed or not
When I study raft, I have a problem.
A Raft cluster has 5 servers. we call them a,b,c,d,e. a is the leader. Now everything is ok.
Then, A handle a client request, makes a log entry.
scenario 1,
b &...
0
votes
1
answer
41
views
Paxos: What happens if the leader lost connection after commit in his own ledger but before multicast the success message?
From the The Part-Time Parliament essay, what happens with the following situation.
The leader receives Voted(b, q) message from the majority and writes this degree
in his own ledger, but lost ...
1
vote
1
answer
410
views
Performance of LWT in scylla/cassandra
Wanted to understand the why do we say the throughput of whole Cassandra/scylla/cockroachDB setup is dependent on LWT/linearised TX.
LWT will affect throughput of a hot key (index or row with a lot of ...
0
votes
1
answer
53
views
How does raft prevent submitted logs from being overwritten
figure 8 in raft paper
Consider a situation like the figure 8 in raft paper, but in (c), log entry from term 2 has been commited, and s1 crashs, s5 becomes leader, then s5 send append entry rpc to s2, ...
1
vote
1
answer
2k
views
what is the key difference between multipaxos and basic paxos protocol
how is multipaxos different from basic paxos? How does ordering work in multipaxos?
Can someone explain multi-paxos along with the diagram
Tried out going through videos and research papers but ...
1
vote
1
answer
488
views
The relationship between Paxos family and data consistency
Paxos, a kind of consensus algorithm, plays a vital role in distributed database systems. It can be used to make the distributed system chooses the same proposal. Data consistency is a big problem in ...
1
vote
1
answer
165
views
Why does Paxos ensure that consensus is reached and does not change?
I have been reading about single-decree Paxos (primarily looking at Paxos Made Simple) yet am confused about if a consensus among Acceptors is guaranteed to not change after being reached.
According ...
2
votes
2
answers
168
views
In Paxos, why can't we use random backoff to avoid collision?
I understand that the heart of Paxos consensus algorithm is that there is only one "majority" in any given set of nodes, therefore if a proposer gets accepted by a majority, there cannot be ...
1
vote
1
answer
73
views
How to implement Byzantine Single-decree Paxos?
I have implemented Single-decree Paxos, and have shown it not to fail in a Monte Carlo network simulator. I followed "Paxos Made Simple" http://lamport.azurewebsites.net/pubs/paxos-simple....
0
votes
1
answer
157
views
How doesn't Hbase use any consensus algorithm like RAFT or Paxos?
I know replication in HBase is done via append only log files, to maintain the same order of entries as that in master to get replicated to slave's WALs I assume there must be some consensus protocol. ...
0
votes
1
answer
152
views
Multiple Consensus in Simple Paxos
How does a group have multiple consensus at different point of time. Let me try to explain the question with an example. Let us assume that there 10 process, among them 3 are proposers(P0, P1, P2). A ...
1
vote
2
answers
293
views
Isabelle proof of correctness for Raft / Paxos algorithm
Can Isabelle help to prove the correctness of the Raft or Paxos consensus algorithm? Is there a formal proof somewhere?
0
votes
1
answer
122
views
distributed system (RPC + Paxos)
I want to ensure consistency between 3 servers.
It's a consensus problem.
Each server listens to the users and saves the collected data in a local file.
I tried to use Rpc to send the new data to all ...
1
vote
1
answer
390
views
How does Basic Paxos proposer know when to increment roundId / proposal number
Looking at the screenshot from this video, around 27:20
https://www.youtube.com/watch?v=JEpsBg0AO6o
When server S5 sends out the prepare RPC, it uses the roundId (proposal number) 4 and server number ...
8
votes
1
answer
4k
views
Difference between 2PC (2 phase commit) and 2 PL (2 phase locking)
What is the difference between the two? The protocol on the surface looks different, but I would like to understand what is really different between the two and why they are not equivalent.
3
votes
1
answer
672
views
Paxos understanding
I have read the paper Paxos made simple.
And after hard thinking, I come to this conclusion:
The Paxos protocol always guarantees the majority of servers accept the same value at the same turn of the ...
1
vote
1
answer
467
views
In Paxos, assume that acceptors do not change their vote, can it reach a livelock?
So, I’m studying Paxos and the Professor made this question:
Assume that acceptors do not change their vote. In other words, if they vote for value v in round i, they will not send learn messages ...
0
votes
2
answers
203
views
Can multiple values be accepted in a single run?
Consider the following flow :
Proposer prepares a message, gets a promise, sends a proposal with some value, gets it accepted. All fine.
After acceptance, another proposer comes along and prepares a ...
1
vote
1
answer
158
views
paxos: why do ids have to increase monotonically?
This is pseudocode for prepare stage:
Is this ID bigger than any round I have previously received?
If yes
store the ID number, max_id = ID
respond with a PROMISE message
If no
...
1
vote
1
answer
201
views
paxos algorithm - how does the propose stage work?
I am looking at the pseudocode for the PROPOSE stage of the paxos algorithm: https://www.cs.rutgers.edu/~pxk/417/notes/paxos.html
did I receive PROMISE responses from a majority of acceptors?
if yes
...
0
votes
1
answer
49
views
How can I understand "value" in bacis paxos
I am reading Lamport's Paxos Mode Simple, and I get confused with the meaning "value" here.
For example, Lamport says:
If a proposal with value v is chosen, then every higher-numbered ...
1
vote
2
answers
317
views
Do proposal numbers need to be unique in Paxos?
What can happen if two proposals are issued with the same ProposalId? It shouldn't create any problem if they have the same value. But what about different values? I can work out a scenario where it ...
-1
votes
1
answer
1k
views
does anyone have any recommendatation for Multi-Paxos? [closed]
I've quite understood what the Raft is and implemented it in MIT6.824 distributed system. I also know what's the basic Paxos, I've not implemented this yet, so I can't grab all details of it. For ...
1
vote
1
answer
239
views
How a distributed storage system like Raft filter duplicate requests even after client session expiration
I am trying to understand how a distributed storage system built on Raft filter duplicate requests even after client session expiration.
I have gone through the Raft dissertation chapter 6.3 which ...
1
vote
0
answers
114
views
Mencius and AllConcur Consensus Algorithms
Just wanted to ask if anyone has had any thoughts on the similarities and differences between the Mencius and AllConcur consensus protocols?
At first glance these algorithms are presented very ...
0
votes
2
answers
674
views
Two Phase Commit blocking on coordinator failure
I was trying to read the Paxos Commit paper and am struggling witih moving past the introduction. The intial section builds a motivation for a fault-tolerant transaction coordinator implementation in ...
0
votes
1
answer
1k
views
Zookeeper zookeeper.forceSync, Zab and Paxos
I noticed that the default the configuration for zookeeper.forceSync is "no".
This mean FileChannel.Force() will not be called on the Write ahead log.
I was under the impression that for the Zab ...
0
votes
1
answer
184
views
Paxos made simple choosing proposing value
The paper says:
1.A proposer chooses a new proposal number n and sends a request to each member of some set of acceptors, asking it to respond with:
(a) A promise never again to accept a ...
3
votes
2
answers
3k
views
Simple leader election (Stateless leader election)
I am building an app in golang that I would like to be fault-tolerant. I looked at different algorithms like RAFT and Paxos and their implementations in golang (etcd's raft, hashicorp's raft), but I ...
0
votes
2
answers
334
views
In paxos, what happens if a proposer is down after its proposal is rejected?
In this figure, the proposal of X is rejected.
At the end of the timeline, S1 and S2 accept X while S3, S4 and S5 accept Y. Proposer X is now supposed to re-send the proposal with value Y.
But what ...
2
votes
1
answer
936
views
Does paxos provide true linearizable consistency or not?
I think I might be confusing concepts here, but it seems to me like paxos would provide linearizable consistency for systems that implement it.
I know Cassandra uses it. I'm not 100% clear on how but ...
3
votes
1
answer
323
views
What is the purpose of Chubby Sequencers
While reading article from google about chubby, I didn't really understand the purpose of sequencers
Assume we have 4 entities :
Chubby cell
Client 1
Client 2
Service we want to use and where we ...
3
votes
2
answers
374
views
how to prove a consensus implementation like multipaxos is right?
I want to prove that my implementation of multi-paxos is right. Are there any valid examples for me to test on? Or there can be some other ways to convince others that my implementation is right.
I ...
5
votes
1
answer
5k
views
Understanding Cassandra Paxos implementation
In Datastax's docs they say that there are four stages in the Paxos protocol (meaning, in a lightweight transaction):
Prepare/Promise
Read/Results
Propose/Accept
Commit/Acknowledge
...
3
votes
1
answer
985
views
Is Paxos Strongly Consistent?
Consider a distributed system with 3 nodes- n1, n2, n3. There is a shared data, x, among the nodes. Paxos is running on the nodes. In the beginning, x is equal to 4.
A client sends an update request ...
0
votes
2
answers
157
views
How a proposer know its propose is not approved by a quorum of acceptors?
I am reading "paxos" on wiki, and it reads:
"Rounds fail when multiple Proposers send conflicting Prepare messages, or when the Proposer does not receive a Quorum of responses (Promise or Accepted). ...
4
votes
1
answer
3k
views
consensus algorithm: what will happen if an odd cluster becomes even because of a node failure?
Consensus algorithm (e.g. raft) requires the cluster contains an odd number of nodes to avoid the split-brain problem.
Say I have a cluster of 5 nodes, what will happen if only one node fails? The ...
5
votes
3
answers
556
views
How do replicas coming back online in PAXOS or RAFT catch up?
In consensus algorithms like for example PAXOS and RAFT, a value is proposed, and if a quorum agrees, it's written durably to the data store. What happens to the participants that were unavailable at ...
1
vote
1
answer
113
views
what decides the roles in Paxos?
When people describe Paxos, they always assume that there are already some proposers in the cluster. But where are the proposers from, or what decides which processes to be proposers?
2
votes
1
answer
119
views
paxos for system builder
Why a separate phase of Leader Election is required in the paper Paxos for System Builders: An Overview instead of using the prepare phase for the leader election? What advantages does this additional ...
3
votes
2
answers
525
views
Commit Failure in Paxos
I am new to the Distributed System and Consensus Algorithm. I understand how it works but I am confused by some corner cases: when the acceptors received an ACCEPT for an instance but never heard back ...
2
votes
2
answers
2k
views
Why Paxos is required for leader election in cassandra
I was reading about the Protocol for distributed consensus and came across the Paxos. It states that you require Paxos for the leader election. Article claiming the same. So, my question is when ...
1
vote
2
answers
417
views
Can paxos consensus be followed in blockchain
Paxos is a fault tolerant distributed consensus algorithm. Are there any blockchain frameworks which follows paxos algorithm for adding blocks to the blockchain.
2
votes
1
answer
358
views
what's the difference between single paxos state machine and multi paxos state machine in spanner paper
From the spanner paper, it says that
"To support replication, each spanserver implements a single Paxos state machine on top of each tablet. (An early Spanner incarnation supported multiple Paxos ...