State machine replication

State machine replication

:"Introduction from Schneider's 1990 survey:":"Distributed software is often structured in terms of clients and services. Each service comprises one or more servers and exports operations that clients invoke by making requests. Although using a single, centralized, server is the simplest way to implement a service, the resulting service can only be as fault tolerant as the processor executing that server. If this level of fault tolerance is unacceptable, then multiple servers that fail independently must be used. Usually, replicas of a single server are executed on separate processors of a distributed system, and protocols are used to coordinate client interactions with these replicas. The physical and electrical isolation of processors in a distributed system ensures that server failures are independent, as required.

:"The state machine approach is a general method for implementing a fault-tolerant service by replicating servers and coordinating client interactions with server replicas. The approach also provides a framework for understanding and designing replication management protocols. Many protocols that involve replication of data or software - be it for masking failures or simply to facilitate cooperation without centralized control - can be derived using the state machine approach. Although few of these protocols actually were obtained in this manner, viewing them in terms of state machines helps in understanding how and why they work." cite journal|curly=yes|last=Schneider|first= Fred|year=1990| title=Implementing Fault-Tolerant Services Using the State Machine Approach: A Tutorial|journal=ACM Computing Surveys|volume=22|url=| doi=10.1145/98163.98167| pages=299| format=Dead link|date=May 2008]


tate Machine Definition

For the subsequent discussion a State Machine will be defined as the following tuple of values cite journal|curly=yes|last=Lamport|first=Leslie|year=1978|title=The Implementation of Reliable Distributed Multiprocess Systems|journal=Computer Networks|volume=2|pages=95–114|url=|accessdate=2008-03-13] Mealy machine:

* A set of States
* A set of Inputs
* A set of Outputs
* A transition function (Input x State -> State)
* An output function (Input x State -> Output)
* A distinguished State called Start.

A State Machine begins at the State labeled Start. Each Input received is passed through the transition and output function to produce a new State and an Output. The State is held stable until a new Input is received, while the Output is communicated to the appropriate receiver.

It should be clear that any algorithm can be implemented using this model if driven by an appropriate Input stream. In particular, this discussion requires a State Machine to have the following property:

:Deterministic::Multiple copies of the same State Machine begun in the Start state, and receiving the same Inputs in the same order will arrive at the same State having generated the same Outputs.

Fault Tolerance Explained

Determinism is an ideal characteristic for providing fault-tolerance. Intuitively, if multiple copies of a system exist, a fault in one would be noticeable as a difference in the State or Output from the others.

A little deduction shows the minimum number of copies needed for fault-tolerance is three; one which has a fault, and two others to whom we compare State and Output. Two copies is not enough; there is no way to tell which copy is the faulty one.

Further deduction shows a three-copy system can support at most one failure (after which it must repair or replace the faulty copy). If more than one of the copies were to fail, all three States and Outputs might differ, and there would be no way to choose which is the correct one.

Research has shown cite web|curly=yes|last=Lamport|first=Leslie|year=2004|title=Lower Bounds for Asynchronous Consensus|url=] that in general a system which supports F failures must have 2F+1 copies (also called replicas). The extra copies are used as evidence to decide which of the copies are correct and which are faulty. Special cases can improve these bounds cite journal|curly=yes|last=Lamport|first=Leslie|coauthors=Mike Massa|year=2004|title=Cheap Paxos|journal=Proceedings of the International Conference on Dependable Systems and Networks (DSN 2004)|url=] .

All of this deduction pre-supposes that replicas are experiencing only random independent faults such as memory errors or hard-drive crash. Failures caused by replicas which attempt to lie, deceive, or collude can also be handled by the State Machine Approach, with isolated changes.

All of this deduction pre-supposes that there is a single value which is the Output of each replica. Faults where a replica sends different values in different directions (for instance, the correct Output to some of its fellow replicas and incorrect Outputs to others) are called Byzantine Failures cite journal|curly=yes|last=Lamport|first=Leslie|coauthors=Robert Shostak, Marshall Pease|year=1982|month=July|title=The Byzantine Generals Problem|journal=ACM Transactions on Programming Languages and Systems|volume=4|issue=3|pages=382–401|url=|accessdate=2007-02-02|doi=10.1145/357172.357176] . Both normal and Byzantine failures are supported by the State Machine Approach, with some changes, but the possibility of Byzantine failures raises the required number of replicas to 3F+1.

It should be noted that failed replicas are not required to stop; they may continue operating, including generating spurious or incorrect Outputs. If all failed replicas are guaranted to stop, we only need F+1 replicas and the system is much simpler: we simply accept the first output that's calculated, since it must've been produced by a replica which hasn't stopped and therefore (by the guarantee) hasn't failed.

The State Machine Approach

The preceding intuitive discussion implies a simple technique for implementing a fault-tolerant service in terms of a State Machine:

# Place copies of the State Machine on multiple, independent servers.
# Receive client requests, interpreted as Inputs to the State Machine.
# Choose an ordering for the Inputs.
# Execute Inputs in the chosen order on each server.
# Respond to clients with the Output from the State Machine.
# Monitor replicas for differences in State or Output.

The remainder of this article develops the details of this technique.

* Step 1 and 2 are outside the scope of this article.
* Step 3 is the critical operation, see Ordering Inputs.
* Step 4 is covered by the State Machine Definition.
* Step 5, see Ordering Outputs.
* Step 6, see Auditing and Failure Detection.

The appendix contains discussion on typical extensions used in real-world systems such as Logging, Checkpoints, Reconfiguration, and State Transfer.

Ordering Inputs

The critical step in building a distributed system of State Machines is choosing an order for the Inputs to be processed. Since all non-faulty replicas will arrive at the same State and Output if given the same Inputs, it is imperative that the Inputs are submitted in an equivalent order at each replica. Many solutions have been proposed in the literature cite journal|curly=yes|last=Lamport|first=Leslie|year=1984|title=Using Time Instead of Timeout for Fault-Tolerant Distributed Systems|journal=ACM Transactions on Programming Languages and Systems|volume=6|issue=2|pages=254–280|url=|accessdate=2008-03-13|doi=10.1145/2993.2994] cite journal|curly=yes|last=Birman|first=Kenneth|coauthors=Thomas Joseph|year=1987|title=Exploiting virtual synchrony in distributed systems|journal= Proceedings of the 11th ACM Symposium on Operating systems principles (SOSP)|url=|accessdate=2008-03-13|doi=10.1145/37499.37515|volume=21|pages=123] cite web|curly=yes|last=Lampson|first=Butler|year=1996|title=How to Build a Highly Available System Using Consensus|url=|accessdate=2008-03-13] .

A Visible Channel is a communication path between two entities actively participating in the system (such as clients and servers). Example: client to server, server to server

A Hidden Channel is a communication path which is not revealed to the system.Example: client to client channels are usually hidden; such as users communicating over a telephone, or a process writing files to disk which are read by another process.

When all communication paths are visible channels and no hidden channels exist, a partial global order (Causal Order) may be inferred from the pattern of communications cite journal|curly=yes|last=Lamport|first=Leslie|year=1978|month=July|title=Time, Clocks and the Ordering of Events in a Distributed System|journal=Communications of the ACM|volume=21|issue=7|pages=558–565|url=|accessdate=2007-02-02|doi=10.1145/359545.359563] . Causal Order may be derived independently by each server. Inputs to the State Machine may be executed in Causal Order, guaranteeing consistent State and Output for all non-faulty replicas.

In open systems, hidden channels are common and a weaker form of ordering must be used. An order of Inputs may be defined using a voting protocol whose results depend only on the visible channels.

The problem of voting for a "single" value by a group of independent entities is called Consensus. By extension, a "series" of values may be chosen by a series of consensus instances. This problem becomes difficult when the participants or their communication medium may experience failures cite web|curly=yes|last=Lamport|first=Leslie|year=2004|title=Lower Bounds for Asynchronous Consensus|url=] .

Inputs may be ordered by their position in the series of consensus instances (Consensus Order). Consensus Order may be derived independently by each server. Inputs to the State Machine may be executed in Consensus Order, guaranteeing consistent State and Output for all non-faulty replicas.

:Optimizing Causal & Consensus Ordering:In some cases additional information is available (such as real-time clocks). In these cases, it is possible to achieve more efficient causal or consensus ordering for the Inputs, with a reduced number of messages, fewer message rounds, or smaller message sizes. See references for details cite web|curly=yes|last=Lamport|first=Leslie|year=2005|title=Fast Paxos|url=]

:Further optimizations are available when the semantics of State Machine operations are accounted for (such as Read vs Write operations). See references Generalized Paxoscite journal|curly=yes|last=Lamport|first=Leslie|year=2005|title=Generalized Consensus and Paxos|url=] .

ending Outputs

Client requests are interpreted as Inputs to the State Machine, and processed into Outputs in the appropriate order. Each replica will generate an Output independently. Non-faulty replicas will always produce the same Output. Before the client response can be sent, faulty Outputs must be filtered out. Typically, a majority of the Replicas will return the same Output, and this Output is sent as the response to the client.

ystem Failure

:If there is no majority of replicas with the same Output, or if less than a majority of replicas returns an Output, a system failure has occurred. The client response must be the unique Output: FAIL.

Auditing and Failure Detection

The permanent, unplanned compromise of a replica is called a Failure. Proof of failure is difficult to obtain, as the replica may simply be slow to respond cite journal|curly=yes|last=Fischer|first=Michael|year=1985|title=Impossibility of Distributed Consensus with One FaultyProcess|journal=Journal of the Association for Computing Machinery|volume=32|issue=2|pages=347–382|url=|accessdate=2008-03-13] , or even lie about its status .

Non-faulty replicas will always contain the same State and produce the same Outputs. This invariant enables failure detection by comparing States and Outputs of all replicas. Typically, a replica with State or Output which differs from the majority of replicas is declared faulty.

A common implementation is to pass checksums of the current replica State and recent Outputs among servers. An Audit process at each server restarts the local replica if a deviation is detected cite journal|curly=yes|last=Chandra|first=Tushar|coauthors=Robert Griesemer, Joshua Redstone|year=2007|title=Paxos Made Live – An Engineering Perspective|journal=PODC '07: 26th ACM Symposium on Principles of Distributed Computing|url=] . Cryptographic security is not required for checksums.

It is possible that the local server is compromised, or that the Audit process is faulty, and the replica continues to operate incorrectly. This case is handled safely by the Output filter described previously (see Sending Outputs).

Appendix: Extensions

Input Log

In a system with no failures, the Inputs may be discarded after being processed by the State Machine. Realistic deployments must compensate for transient non-failure behaviors of the system such as message loss, network partitions, and slow processors .

One technique is to store the series of Inputs in a log. During times of transient behavior, replicas may request copies of a log entry from another replica in order to fill in missing Inputs .

In general the log is not required to be persistent (it may be held in memory). A persistent log may compensate for extended transient periods, or support additional system features such as Checkpoints, and Reconfiguration.


If left unchecked a log will grow until it exhausts all available storage resources. For continued operation, it is necessary to forget log entries. In general a log entry may be forgotten when its contents are no longer relevant (for instance if all replicas have processed an Input, the knowledge of the Input is no longer needed).

A common technique to control log size is store a duplicate State (called a Checkpoint), then discard any log entries which contributed to the checkpoint. This saves space when the duplicated State is smaller than the size of the log.

Checkpoints may be added to any State Machine by supporting an additional Input called CHECKPOINT. Each replica maintains a checkpoint in addition to the current State value. When the log grows large, a replica submits the CHECKPOINT command just like a client request. The system will ensure non-faulty replicas process this command in the same order, after which all log entries before the checkpoint may be discarded.

In a system with checkpoints, requests for log entries occurring before the checkpoint are ignored. Replicas which cannot locate copies of a needed log entry are faulty and must re-join the system (see Reconfiguration).


Reconfiguration allows replicas to be added and removed from a system while client requests continue to be processed. Planned maintenance and replica failure are common examples of reconfiguration. Reconfiguration involves Quitting and Joining.


When a server detects its State or Output is faulty (see Auditing and Failure Detection), it may selectively exit the system. Likewise, an administrator may manually execute a command to remove a replica for maintenance.

A new Input is added to the State Machine called QUIT. A replica submits this command to the system just like a client request. All non-faulty replicas remove the quitting replica from the system upon processing this Input. During this time, the replica may ignore all protocol messages. If a majority of non-faulty replicas remain, the quit is successful. If not, there is a System Failure.


After quitting, a failed server may selectively restart or re-join the system. Likewise, an administrator may add a new replica to the group for additional capacity.

A new Input is added to the State Machine called JOIN. A replica submits this command to the system just like a client request. All non-faulty replicas add the joining node to the system upon processing this Input. A new replica must be up-to-date on the system's State before joining (see State Transfer).

tate Transfer

When a new replica is made available or an old replica is restarted, it must be brought up to the current State before processing Inputs (see Joining). Logically, this requires applying every Input from the dawn of the system in the appropriate order.

Typical deployments short-circuit the logical flow by performing a State Transfer of the most recent Checkpoint (see Checkpoints). This involves directly copying the State of one replica to another using an out-of-band protocol.

A checkpoint may be large, requiring an extended transfer period. During this time, new Inputs may be added to the log. If this occurs, the new replica must also receive the new Inputs and apply them after the checkpoint is received. Typical deployments add the new replica as an observer to the ordering protocol before beginning the state transfer, allowing the new replica to collect Inputs during this period.

:Optimizing State Transfer:Common deployments reduce state transfer times by sending only State components which differ. This requires knowledge of the State Machine internals. Since state transfer is usually an out-of-band protocol, this assumption is not difficult to achieve.

:Compression is another feature commonly added to state transfer protocols, reducing the size of the total transfer.

Leader Election (for Paxos)

Paxoscite journal|curly=yes|last=Lamport|first=Leslie|year=1998|month=May|title=The Part-Time Parliament|journal=ACM Transactions on Computer Systems|volume=16|issue=2|pages=133–169|url=|accessdate=2007-02-02|doi=10.1145/279227.279229] is a protocol for solving consensus, and may be used as the protocol for implementing Consensus Order.

Paxos requires a single leader to ensure liveness . That is, one of the replicas must remain leader long enough to achieve consensus on the next operation of the state machine. System behavior is unaffected if the leader changes after every instance, or if the leader changes multiple times per instance. The only requirement is that one replica remain leader long enough to move the system forward.

:Conflict Resolution:In general, a leader is necessary only when there is disagreement about which operation to perform , and if those operations conflict in some way (for instance, if they do not commute) .

:When conflicting operations are proposed, the leader acts as the single authority to set the record straight, defining an order for the operations, allowing the system to make progress.

With Paxos, multiple replicas may believe they are leaders at the same time. This property makes Leader Election for Paxos very simple, and any algorithm which guarantees an 'eventual leader' will work.

Historical background

Leslie Lamport was the first to propose the state machine approach, in his seminal 1984 paper on [ "Using Time Instead of Timeout In Distributed Systems"] . Fred Schneider later elaborated the approach in his paper [ "Implementing Fault-Tolerant Services Using the State Machine Approach: A Tutorial"] .

[ 'Ken Birman'] developed the virtual synchrony model in a series of papers published between 1985 and 1987. The primary reference to this work is [ "Exploiting Virtual Synchrony in Distributed Systems"] , which describes the Isis Toolkit, a system that was used to build the New York and Swiss Stock Exchanges, French Air Traffic Control System, US Navy AEGIS Warship, and other applications.

Recent work by Miguel Castro and Barbara Liskov used the state machine approach in what they call a [ "Practical Byzantine fault tolerance"] architecture that replicates especially sensitive services using a version of Lamport's original state machine approach, but with optimizations that substantially improve performance.

A hard real-time variant of this approach has been developed by Prof. Hermann Kopetz at TU Vienna, Austria, in the [ "Time-Triggered Architecture"] (TTA) based on the Time-Triggered Protocol (TTP) during the 1990s. It has been commercialized in the 2000s by TTTech Computertechnik AG and deployed in various aerospace projects.


Wikimedia Foundation. 2010.

Игры ⚽ Нужно решить контрольную?

Look at other dictionaries:

  • Finite-state machine — State machine redirects here. For infinite state machines, see State transition system. For fault tolerance methodology, see State machine replication. SFSM redirects here. For the Italian railway company, see Circumvesuviana. A finite state… …   Wikipedia

  • Replication (computer science) — Replication is the process of sharing information so as to ensure consistency between redundant resources, such as software or hardware components, to improve reliability, fault tolerance, or accessibility. It could be data replication if the… …   Wikipedia

  • Machine de Von Neumann — Constructeur universel Le constructeur universel est une structure autoréplicante au sein d un automate cellulaire, tous deux créés par John von Neumann dans les années 1940[1]. Sommaire 1 Description 2 Interprétation …   Wikipédia en Français

  • Self-replication — is any process by which a thing might make a copy of itself. Biological cells, given suitable environments, reproduce by cell division. During cell division, DNA is replicated and can be transmitted to offspring during reproduction. Biological… …   Wikipedia

  • Virtual synchrony — is an interprocess messaging passing (sometimes called event queue management) technology. Virtual synchrony systems allow programs running in a network to organize themselves into process groups , and to send messages to groups (as opposed to… …   Wikipedia

  • Consensus (computer science) — Consensus is a problem in distributed computing that encapsulates the task of group agreement in the presence of faults.[1] In particular, any process in the group may fail at any time. Consensus is fundamental to core techniques in fault… …   Wikipedia

  • Technical features new to Windows Vista — This article is part of a series on Windows Vista New features Overview Technical and core system Security and safety Networking technologies I/O technologies Management and administration Removed features …   Wikipedia

  • Conway's Game of Life — Conway game , which redirects to here, can also refer to games as defined by surreal numbers, which John Conway also developed …   Wikipedia

  • Paxos algorithm — Paxos is a family of protocols for solving consensus in a network of unreliable processors.Consensus is the process of agreeing on one result among a group of participants. This problem becomes difficult when the participants or their… …   Wikipedia

  • Simulation — Simulator redirects here. For other uses, see Simulator (disambiguation). For other uses, see Simulation (disambiguation). Not to be confused with Stimulation. Wooden mechanical horse simulator during WWI. Simulation is the imitation of some real …   Wikipedia

Share the article and excerpts

Direct link
Do a right-click on the link above
and select “Copy Link”