This document discusses synchronization in distributed systems and various algorithms for achieving mutual exclusion. It covers centralized, distributed, and token ring algorithms for mutual exclusion. The centralized algorithm uses a coordinator but has a single point of failure. Distributed algorithms overcome this but require more messages. The token ring algorithm passes a token between processes but can lose the token if a process crashes. In comparing the algorithms, the document examines their message requirements, delay before entry, and potential problems.

1. Distributed Mutual Exclusion
• Synchronization in Distributed Systems
Synchronization in distributed systems are often more difficult
compared to synchronization in uniprocessor and multiprocessor
systems.

2. Synchronization in Distributed
Systems
 Synchronization in time – achieved by Clock
Synchronization algorithms
 Synchronization between resources - Mutual
Exclusion algorithms
 Synchronization in activities - Distributed
Transaction paradigms

3. Distributed Mutual Exclusion
 Mutual Exclusion
Mutual exclusion (often abbreviated to mutex) algorithms are
used in concurrent programming to avoid the simultaneous use
of a common resource, such as a global variable, by pieces
of computer code called critical sections.

4. Distributed Mutual Exclusion
 Critical Sections
In concurrent programming a critical section is a piece of code
that accesses a shared resource (data structure or device) that
must not be concurrently accessed by more than one thread of
execution. A critical section will usually terminate in fixed
time, and a thread, task or process will only have to wait a fixed
time to enter it. Some synchronization mechanism is required at
the entry and exit of the critical section to ensure exclusive
use, for example a semaphore.

5. Distributed Mutual Exclusion
 Achieving Mutual exclusion
 Hardware solutions - Disabling interrupts on entry into the
critical section
 System variables - By using semaphores implemented as
Locks
 Software solutions - Mutual Exclusion Algorithms

6. Distributed Mutual Exclusion
 Mutual Exclusion Algorithms
 Centralized algorithms
 Distributed algorithms
 Token ring algorithm

7. Mutual Exclusion – centralized Algorithm
a) Process 1 asks the coordinator for permission to enter a critical
region. Permission is granted
b) Process 2 then asks permission to enter the same critical
region. The coordinator does not reply.
c) When process 1 exits the critical region, it tells the
coordinator, when then replies to 2

8. Centralized Algorithm – Advantages Vs
Disadvantages
 Advantages
 Fair algorithm, grants in the order of requests
 The scheme is easy to implement
 Scheme can be used for general resource allocation
 Shortcomings
 Single point of failure. No fault tolerance
 Confusion between No-reply and permission denied
 Performance bottleneck of single co-ordinator in a large
system

9. Mutual Exclusion – Distributed Algorithm
a) Two processes want to enter the same critical region at the
same moment.
b) Process 0 has the lowest timestamp, so it wins.
c) When process 0 is done, it sends an OK also, so 2 can now
enter the critical region.

10. Distributed Algorithm – Advantages Vs
Disadvantages
 Advantages
 Single point of failure is overcome by keeping the system
running even if one processes fail
 No bottleneck on one process
 Shortcomings
 2(n-1) messages or more than that
 Waiting for n-1 processes to accept is overkill
 Improvements
 Group membership for easy communication

11. A Token Ring Algorithm
a) An unordered group of processes on a
network.
b) A logical ring constructed in software.

12. Token Ring Algorithm – Advantages Vs
Disadvantages
 Advantages
 No co-ordinator and does not depend on other
processes
 On Crash token is just passed on to the neighbour
 Shortcomings
 Only a physical Q, no logical Q
 Loss of token – Regeneration of tokens
 Difficulty in finding the token crash – The
processor might be alive

13. Comparison
 A comparison of three mutual exclusion algorithms.
Algorithm
Messages per
entry/exit
Delay before entry
(in message times)
Problems
Centralized 3 2 Coordinator crash
Distributed 2 ( n – 1 ) 2 ( n – 1 ) Overkill in waiting
Token ring 1 to 0 to n – 1
Lost token on
process crash

14. References
 Distributed systems – Principles and Paradigms
– Andrew S Tanenbaum
 A Survey of Mutual-Exclusion Algorithms for
Multiprocessor Operating Systems
http://www.teamten.com/lawrence/242.paper/242.paper.html