View
237
Download
0
Embed Size (px)
Citation preview
Chapter 18-1: Distributed Chapter 18-1: Distributed CoordinationCoordination
18.2 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Chapter 18 Distributed CoordinationChapter 18 Distributed Coordination
Chapter 18.1
Event Ordering
Mutual Exclusion
Atomicity
Chapter 18.2
Concurrency Control
Deadlock Handling
Election Algorithms
Reaching Agreement
18.3 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Chapter ObjectivesChapter Objectives
Chapter 18.1 To describe various methods for achieving mutual
exclusion in a distributed system To explain how atomic transactions can be
implemented in a distributed system
Chapter 18.2 To show how some of the concurrency-control schemes
discussed in Chapter 6 can be modified for use in a distributed environment
To present schemes for handling deadlock prevention, deadlock avoidance, and deadlock detection in a distributed system
18.4 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
BackgroundBackground
When we are dealing with distributed systems, we have unique activities that must be addressed.
A number of these deal with mutual exclusion and synchronization.
Is synchronization handled centrally or not?
Problems with deadlock are clearly related.
Many of the activities – and solutions – deal with an ordering of events that take place.
So we will start out with Event Ordering.
18.5 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
1. Event Ordering1. Event Ordering
Ordering in a centralized system is facilitated with common memory and a single clock.
We can ensure that one event ‘occurs’ before another.
We know, for example, that we can allocate a resource to a process after a different process releases it.
We can say that one event must occur prior to a second event.
Clearly, in a distributed environment, this is not an easy case…
18.6 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Event Ordering - continuedEvent Ordering - continued
In a distributed system, we need a distributed algorithm to control a ‘happened before’ relation and apply this to all events in the system.
Notation: Let’s use the following conventions: as applied to sequential processes:
Happened-before relation (denoted by ) If A and B are events in the same process, and A was executed before B,
then A B If A is the event of sending a message by one process and B is the event
of receiving that message by another process, then A B If A B and B C then A C Further, an event cannot occur before itself.
Concurrent Events: If two events are NOT related by , that is A did not happen before B and B did not happen before A, we say that these two events were executed concurrently.
There is no cause-effect relationship between the two; however, it is possible for event A to affect event B causally.
Let’s consider the next slide…
18.7 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Relative Time for Three Concurrent ProcessesRelative Time for Three Concurrent Processes
Labeled dots represent events that are part of processes (or processors).The wavy line implies that a message is sent from one process to the other.Events are concurrent “iff” no path exists between them.
Can see p1 q2; p1 q4 (since p1q2 and q2 q4)
But there are some concurrent events here too: q0 and p2; r0 and q3 and others.Now, importantly, we don’t know which of q0 and p2 occurred first.But for happened before events, there is some kind of order agreed to.
18.8 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Implementation of Implementation of To implement a ‘happened before’ feature, we need a common clock or a
perfectly synchronized set of clocks. But we don’t have these in distributed systems. One approach is to use a timestamp for each event. Then we can establish some kind of global ordering By require this for all system events, then for every pair of events A and B,
if A B, then the timestamp of A is less than the timestamp of B
But how do we do this? Within each process Pi we have a logical clock, LCi associated with it.
The logical clock can be implemented as a simple counter incremented in a monotonically increasing fashion. Thus a number can be assigned to each event and these values can be compared.
So, for events within the same process, this notion of a global ordering is met, and this is good, But…this approach does not ensure a global ordering across different processes / processors..
18.9 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Implementation of Implementation of
For different processors:
Issues arise almost immediately with the recognition that clocks on certain processors operate more slowly than for others.
So what to do?
One approach is to require a process to advance its logical clock when it receives a message whose timestamp is greater than the current value of its clock.
This makes perfect sense and thus puts the ‘logical clocks’ compatible.
Only fly in the ointment is if two timestamps are the same.
Here, events are considered concurrent.
There are other techniques to deal with this – ahead. In particular, we can use process identity numbers to break ties and to create the total ordering we need.
18.10 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
2. Distributed Mutual Exclusion (DME) 2. Distributed Mutual Exclusion (DME)
Here, we will discuss mutual exclusion, deadlock, starvation problems that may occur in a distributed system.
Assumptions The system consists of n processes; each process Pi resides at a
different processor Each process has a critical section that requires mutual exclusion Processes are numbered sequentially from 1 to n; result is that each
process has its own processor. Requirement
If Pi is executing in its critical section, then no other process Pj is executing in its critical section
We present two algorithms to ensure the mutual exclusion execution of processes in their critical sections
The two approaches are the Centralized Approach and the Fully-Distributed Approach
18.11 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
DME: Centralized Approach (1 of 2)DME: Centralized Approach (1 of 2)
This is a simpler approach by far: One of the processes in the system is chosen to coordinate the entry
to the critical section A process that wants to enter its critical section sends a request
message to the coordinator The coordinator decides which process can enter the critical section
next, and its sends that process a reply message When the process receives a reply message from the coordinator, it
enters its critical section After exiting its critical section, the process sends a release message
to the coordinator and proceeds with its execution This scheme requires three messages per critical-section entry:
request reply release
18.12 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
DME: Centralized Approach (2 of 2)DME: Centralized Approach (2 of 2)
The coordinator must determine if any other process is in its critical section in order to send back a reply message. If a process is in its critical section, then the request is queued. When the coordinator receives a release message, it can access
its queue of queued request messages and then send a reply message to the initiating process.
This implementation is essentially a FCFS scheduler. Thus it is: Fair, and Prevents Starvation, and Ensures mutual exclusion.
Of course, if the coordinator fails then there must be some algorithm available to reestablish a coordinator which will poll processes in the system in order to rebuilt its request queue.
18.13 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
DME: Fully Distributed ApproachDME: Fully Distributed Approach
Ah, but now we have the real power …. And complexity!!
When process Pi wants to enter its critical section, it generates a new timestamp, TS, and sends the message request (Pi, TS) to all other processes in the system
When process Pj receives a request message, it may
reply immediately or it
may defer sending a reply back
When process Pi receives a reply message from all other processes in the system, it can enter its critical section
After exiting its critical section, the process sends reply messages to all its deferred requests
18.14 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
DME: Fully Distributed Approach (Cont.)DME: Fully Distributed Approach (Cont.)
The decision whether process Pj replies immediately to a request(Pi, TS) message or defers its reply is based on three factors:
If Pj is in its critical section, then it defers its reply to Pi
If Pj does not want to enter its critical section, then it sends a reply immediately to Pi
If Pj wants to enter its critical section but has not yet entered it, then it compares its own request timestamp with the timestamp TS
If its own request timestamp is greater than TS, then it sends a reply immediately to Pi (Pi asked first)
Otherwise, the reply is deferred
18.15 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Desirable Behavior of Fully Distributed ApproachDesirable Behavior of Fully Distributed Approach
This distributed approach provides three nice features:
Mutual exclusion is established – as before,
Freedom from Deadlock is ensured
Freedom from starvation is ensured, since entry to the critical section is scheduled according to the timestamp ordering and all requests will sometime get a reply message returned.
The timestamp ordering ensures that processes are served in a first-come, first served order
So, these look pretty good. But unfortunately, there are three ‘downsides.’
18.16 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Three Undesirable ConsequencesThree Undesirable Consequences
1. The processes need to know the identity of all other processes in the system, which makes the dynamic addition and removal of processes more complex
This means that processes recdeive names of other processes and the names of a new process must be distributed to all other processes in the group!
If one of the processes fails, then the entire scheme collapses This can be dealt with by continuously monitoring the state of all the processes in the
system If a process goes down, then all processes need to be notified so that they don’t send
request messages to that process. Of course, when a process is restored, all processes must be similarly informed so that they can send request messages to that restored site.
Clearly this introduces a great deal of overhead!
Processes that have not entered their critical section must pause frequently to assure other processes that they intend to enter the critical section
This protocol is therefore suited for small, stable sets of cooperating processes
18.17 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Token-Passing ApproachToken-Passing Approach
Circulate a token among processes in system
Token is special type of message
Possession of token entitles holder to enter critical section
Processes logically organized in a ring structure
Don’t need to literally be a ring, but organized like a ring so that a token can be passed from one process to another.
Once process is complete, the token is passed to the next process in the logical ring.
Unidirectional ring guarantees freedom from starvation
Two types of failures
Lost token – election must be called
Failed processes – new logical ring established and all processes must be informed.
All must be informed with a process is restored as well.
18.18 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Atomicity Atomicity
Either all the operations associated with a program unit are executed to completion, or none are performed
This is a complicated process in a distributed system.
Ensuring atomicity in a distributed system requires a transaction coordinator, which is responsible for the following:
Starting the execution of the transaction
Breaking the transaction into a number of subtransactions, and distribution these subtransactions to the appropriate sites for execution
Coordinating the termination of the transaction, which may result in the transaction being committed at all sites or aborted at all sites
Also note that each site has its own transaction coordinator for requests sent to it. It, in turn, must coordinate the execution of all transactions initiated at that site.
18.19 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Two-Phase Commit Protocol (2PC)Two-Phase Commit Protocol (2PC)
This is pretty interesting and complicated…
Since we are talking distributed execution, parts of a process may be executed at a number of cooperating sites.
For atomicity to be guaranteed, all participating sites must be on board.
A transaction T must either commit at all sites or must abort at all sites. There is no middle ground.
To guarantee all sites are on board – more precisely, to guarantee atomicity, the transaction coordinator of T must execute a commit protocol, the most popular of which is a two-phase commit (2PC) protocol.
Consider a transaction T, at site S that has a transaction coordinator C at that site.
When all sites where T (or parts of T) has been executed tell the coordinator, C, that T is done, the coordinator, C, starts the 2PC protocol.
18.20 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Phase 1: Obtaining a DecisionPhase 1: Obtaining a Decision
Ci adds <prepare T> record to the log and records this record on stable storage.
Ci sends <prepare T> message to all sites.
When a site receives a <prepare T> message, the transaction manager at that site determines if it can commit the transaction.
If this transaction manager decides on ‘no’: it adds a <no T> record to the log and responds to Ci with <abort T> message.
If yes, the transaction manager at this site:
adds a <ready T> record to the log
force all log records for T onto stable storage
transaction manager sends <ready T> message to Ci
18.21 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Phase 1 (Cont.)Phase 1 (Cont.)
The initial Transaction Coordinator collects responses
All respond “ready”, decision is commit
At least one response is “abort”,decision is abort
At least one participant fails to respond within time out period,decision is abort
18.22 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Phase 2: Recording Decision in the DatabasePhase 2: Recording Decision in the Database
Then the Transaction Coordinator adds a decision record
<abort T> or <commit T>
to its log and forces record onto stable storage
Once that record reaches stable storage it is irrevocable (even if failures occur)
Coordinator sends a message to each participant informing it of the decision (commit or abort)
Receiving site records the message in the log; that is, participants take appropriate action locally
18.23 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
More on 2PCMore on 2PC It is interesting to note that any site that T or part of T has be executed may
unconditionally abort T at any time before sending a ready T message to the coordinator.
The ready T is essentially a promise to follow the coordinator’s order to commit T or to abort T.
But the site can only make such a promise when all the needed information it requires is available in its stable storage (disk) so that in the event the site crashes, the non-volatility nature of the stable-storage enables it to follow through on its promise, once it is up and running again.
Also interesting that it only takes a single abort T to kill any potential commit T.
But the final decision is rendered by the transaction coordinator which writes this decision to its log and forces it onto stable storage.
A nice ‘touch’ in some implementations of the 2PC protocol is that some sites send an acknowledge T back to the coordinator at the end of the second phase of the 2PC protocol.
18.24 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Failure Handling in 2PC – Site FailureFailure Handling in 2PC – Site Failure
The log contains a <commit T> record
In this case, the site executes redo(T)
The log contains an <abort T> record
In this case, the site executes undo(T)
The contains a <ready T> record; consult Ci
If Ci is down, site sends query-status T message to the other sites
The log contains no control records concerning T
In this case, the site executes undo(T)
18.25 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Failure Handling in 2PC – Coordinator Failure Handling in 2PC – Coordinator CCii FailureFailure
If an active site contains a <commit T> record in its log, the T must be committed
If an active site contains an <abort T> record in its log, then T must be aborted
If some active site does not contain the record <ready T> in its log then the failed coordinator Ci cannot have decided to commit T
Rather than wait for Ci to recover, it is preferable to abort T
All active sites have a <ready T> record in their logs, but no additional control records
In this case we must wait for the coordinator to recover
Blocking problem – T is blocked pending the recovery of site Si
18.26 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Concurrency ControlConcurrency Control
Modify the centralized concurrency schemes to accommodate the distribution of transactions
Transaction manager coordinates execution of transactions (or subtransactions) that access data at local sites
Local transaction only executes at that site
Global transaction executes at several sites
18.27 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Locking ProtocolsLocking Protocols
Can use the two-phase locking protocol in a distributed environment by changing how the lock manager is implemented
Nonreplicated scheme – each site maintains a local lock manager which administers lock and unlock requests for those data items that are stored in that site
Simple implementation involves two message transfers for handling lock requests, and one message transfer for handling unlock requests
Deadlock handling is more complex
18.28 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Single-Coordinator ApproachSingle-Coordinator Approach
A single lock manager resides in a single chosen site, all lock and unlock requests are made a that site
Simple implementation
Simple deadlock handling
Possibility of bottleneck
Vulnerable to loss of concurrency controller if single site fails
Multiple-coordinator approach distributes lock-manager function over several sites
18.29 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Majority ProtocolMajority Protocol
Avoids drawbacks of central control by dealing with replicated data in a decentralized manner
More complicated to implement
Deadlock-handling algorithms must be modified; possible for deadlock to occur in locking only one data item
18.30 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Biased ProtocolBiased Protocol
Similar to majority protocol, but requests for shared locks prioritized over requests for exclusive locks
Less overhead on read operations than in majority protocol; but has additional overhead on writes
Like majority protocol, deadlock handling is complex
18.31 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Primary CopyPrimary Copy
One of the sites at which a replica resides is designated as the primary site
Request to lock a data item is made at the primary site of that data item
Concurrency control for replicated data handled in a manner similar to that of unreplicated data
Simple implementation, but if primary site fails, the data item is unavailable, even though other sites may have a replica
18.32 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
TimestampingTimestamping
Generate unique timestamps in distributed scheme:
Each site generates a unique local timestamp
The global unique timestamp is obtained by concatenation of the unique local timestamp with the unique site identifier
Use a logical clock defined within each site to ensure the fair generation of timestamps
Timestamp-ordering scheme – combine the centralized concurrency control timestamp scheme with the 2PC protocol to obtain a protocol that ensures serializability with no cascading rollbacks
18.33 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Generation of Unique TimestampsGeneration of Unique Timestamps
18.34 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Deadlock PreventionDeadlock Prevention
Resource-ordering deadlock-prevention – define a global ordering among the system resources
Assign a unique number to all system resources
A process may request a resource with unique number i only if it is not holding a resource with a unique number grater than i
Simple to implement; requires little overhead
Banker’s algorithm – designate one of the processes in the system as the process that maintains the information necessary to carry out the Banker’s algorithm
Also implemented easily, but may require too much overhead
18.35 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Timestamped Deadlock-Prevention SchemeTimestamped Deadlock-Prevention Scheme
Each process Pi is assigned a unique priority number
Priority numbers are used to decide whether a process Pi should wait for a process Pj; otherwise Pi is rolled back
The scheme prevents deadlocks
For every edge Pi Pj in the wait-for graph, Pi has a higher priority than Pj
Thus a cycle cannot exist
18.36 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Wait-Die SchemeWait-Die Scheme
Based on a nonpreemptive technique
If Pi requests a resource currently held by Pj, Pi is allowed to wait only if it has a smaller timestamp than does Pj (Pi is older than Pj)
Otherwise, Pi is rolled back (dies)
Example: Suppose that processes P1, P2, and P3 have timestamps t, 10, and 15 respectively
if P1 request a resource held by P2, then P1 will wait
If P3 requests a resource held by P2, then P3 will be rolled back
18.37 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Would-Wait SchemeWould-Wait Scheme
Based on a preemptive technique; counterpart to the wait-die system
If Pi requests a resource currently held by Pj, Pi is allowed to wait only if it has a larger timestamp than does Pj (Pi is younger than Pj). Otherwise Pj is rolled back (Pj is wounded by Pi)
Example: Suppose that processes P1, P2, and P3 have timestamps 5, 10, and 15 respectively
If P1 requests a resource held by P2, then the resource will be preempted from P2 and P2 will be rolled back
If P3 requests a resource held by P2, then P3 will wait
18.38 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Two Local Wait-For GraphsTwo Local Wait-For Graphs
18.39 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Global Wait-For GraphGlobal Wait-For Graph
18.40 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Deadlock Detection – Centralized ApproachDeadlock Detection – Centralized Approach
Each site keeps a local wait-for graph
The nodes of the graph correspond to all the processes that are currently either holding or requesting any of the resources local to that site
A global wait-for graph is maintained in a single coordination process; this graph is the union of all local wait-for graphs
There are three different options (points in time) when the wait-for graph may be constructed:
1. Whenever a new edge is inserted or removed in one of the local wait-for graphs
2. Periodically, when a number of changes have occurred in a wait-for graph
3. Whenever the coordinator needs to invoke the cycle-detection algorithm
Unnecessary rollbacks may occur as a result of false cycles
18.41 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Detection Algorithm Based on Option 3Detection Algorithm Based on Option 3
Append unique identifiers (timestamps) to requests form different sites
When process Pi, at site A, requests a resource from process Pj, at site B, a request message with timestamp TS is sent
The edge Pi Pj with the label TS is inserted in the local wait-for of A. The edge is inserted in the local wait-for graph of B only if B has received the request message and cannot immediately grant the requested resource
18.42 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
The Algorithm The Algorithm
1. The controller sends an initiating message to each site in the system
2. On receiving this message, a site sends its local wait-for graph to the coordinator
3. When the controller has received a reply from each site, it constructs a graph as follows:
(a) The constructed graph contains a vertex for every process in the system
(b) The graph has an edge Pi Pj if and only if
(1) there is an edge Pi Pj in one of the wait-for graphs, or
(2) an edge Pi Pj with some label TS appears in more than one wait-for graph
If the constructed graph contains a cycle deadlock
18.43 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Local and Global Wait-For GraphsLocal and Global Wait-For Graphs
18.44 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Fully Distributed ApproachFully Distributed Approach
All controllers share equally the responsibility for detecting deadlock
Every site constructs a wait-for graph that represents a part of the total graph
We add one additional node Pex to each local wait-for graph
If a local wait-for graph contains a cycle that does not involve node Pex, then the system is in a deadlock state
A cycle involving Pex implies the possibility of a deadlock
To ascertain whether a deadlock does exist, a distributed deadlock-detection algorithm must be invoked
18.45 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Augmented Local Wait-For Graphs Augmented Local Wait-For Graphs
18.46 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Augmented Local Wait-For Graph in Site S2Augmented Local Wait-For Graph in Site S2
18.47 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Election AlgorithmsElection Algorithms
Determine where a new copy of the coordinator should be restarted
Assume that a unique priority number is associated with each active process in the system, and assume that the priority number of process Pi is i
Assume a one-to-one correspondence between processes and sites
The coordinator is always the process with the largest priority number. When a coordinator fails, the algorithm must elect that active process with the largest priority number
Two algorithms, the bully algorithm and a ring algorithm, can be used to elect a new coordinator in case of failures
18.48 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Bully AlgorithmBully Algorithm
Applicable to systems where every process can send a message to every other process in the system
If process Pi sends a request that is not answered by the coordinator within a time interval T, assume that the coordinator has failed; Pi tries to elect itself as the new coordinator
Pi sends an election message to every process with a higher priority number, Pi then waits for any of these processes to answer within T
18.49 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Bully Algorithm (Cont.)Bully Algorithm (Cont.)
If no response within T, assume that all processes with numbers greater than i have failed; Pi elects itself the new coordinator
If answer is received, Pi begins time interval T´, waiting to receive a message that a process with a higher priority number has been elected
If no message is sent within T´, assume the process with a higher number has failed; Pi should restart the algorithm
18.50 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Bully Algorithm (Cont.)Bully Algorithm (Cont.)
If Pi is not the coordinator, then, at any time during execution, Pi
may receive one of the following two messages from process Pj
Pj is the new coordinator (j > i). Pi, in turn, records this information
Pj started an election (j > i). Pi, sends a response to Pj and begins its own election algorithm, provided that Pi has not already initiated such an election
After a failed process recovers, it immediately begins execution of the same algorithm
If there are no active processes with higher numbers, the recovered process forces all processes with lower number to let it become the coordinator process, even if there is a currently active coordinator with a lower number
18.51 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Ring AlgorithmRing Algorithm
Applicable to systems organized as a ring (logically or physically)
Assumes that the links are unidirectional, and that processes send their messages to their right neighbors
Each process maintains an active list, consisting of all the priority numbers of all active processes in the system when the algorithm ends
If process Pi detects a coordinator failure, I creates a new active list that is initially empty. It then sends a message elect(i) to its right neighbor, and adds the number i to its active list
18.52 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Ring Algorithm (Cont.)Ring Algorithm (Cont.)
If Pi receives a message elect(j) from the process on the left, it must respond in one of three ways:
1. If this is the first elect message it has seen or sent, Pi creates a new active list with the numbers i and j
It then sends the message elect(i), followed by the message elect(j)
2. If i j, then the active list for Pi now contains the numbers of all the active processes in the system
Pi can now determine the largest number in the active list to identify the new coordinator process
3. If i = j, then Pi receives the message elect(i)
The active list for Pi contains all the active processes in the system
Pi can now determine the new coordinator process.
18.53 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Reaching AgreementReaching Agreement
There are applications where a set of processes wish to agree on a common “value”
Such agreement may not take place due to:
Faulty communication medium
Faulty processes
Processes may send garbled or incorrect messages to other processes
A subset of the processes may collaborate with each other in an attempt to defeat the scheme
18.54 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Faulty CommunicationsFaulty Communications
Process Pi at site A, has sent a message to process Pj at site B; to proceed, Pi needs to know if Pj has received the message
Detect failures using a time-out scheme When Pi sends out a message, it also specifies a time interval
during which it is willing to wait for an acknowledgment message form Pj
When Pj receives the message, it immediately sends an acknowledgment to Pi
If Pi receives the acknowledgment message within the specified time interval, it concludes that Pj has received its message
If a time-out occurs, Pj needs to retransmit its message and wait for an acknowledgment
Continue until Pi either receives an acknowledgment, or is notified by the system that B is down
18.55 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Faulty Communications (Cont.)Faulty Communications (Cont.)
Suppose that Pj also needs to know that Pi has received its acknowledgment message, in order to decide on how to proceed
In the presence of failure, it is not possible to accomplish this task
It is not possible in a distributed environment for processes Pi and Pj to agree completely on their respective states
18.56 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Faulty Processes (Byzantine Generals Problem)Faulty Processes (Byzantine Generals Problem)
Communication medium is reliable, but processes can fail in unpredictable ways
Consider a system of n processes, of which no more than m are faulty
Suppose that each process Pi has some private value of Vi
Devise an algorithm that allows each nonfaulty Pi to construct a vector Xi = (Ai,1, Ai,2, …, Ai,n) such that::
If Pj is a nonfaulty process, then Aij = Vj.
If Pi and Pj are both nonfaulty processes, then Xi = Xj.
Solutions share the following properties
A correct algorithm can be devised only if n 3 x m + 1
The worst-case delay for reaching agreement is proportionate to m + 1 message-passing delays
18.57 Silberschatz, Galvin and Gagne ©2005Operating System Concepts
Faulty Processes (Cont.)Faulty Processes (Cont.)
An algorithm for the case where m = 1 and n = 4 requires two rounds of information exchange:
Each process sends its private value to the other 3 processes
Each process sends the information it has obtained in the first round to all other processes
If a faulty process refuses to send messages, a nonfaulty process can choose an arbitrary value and pretend that that value was sent by that process
After the two rounds are completed, a nonfaulty process Pi can construct its vector Xi = (Ai,1, Ai,2, Ai,3, Ai,4) as follows:
Ai,j = Vi
For j i, if at least two of the three values reported for process Pj agree, then the majority value is used to set the value of Aij
Otherwise, a default value (nil) is used
End of Chapter 18End of Chapter 18