Scribd
Upload
2 views3 pages

Distributed Systems Synchronization Assignment

The document discusses synchronization in distributed systems, highlighting the importance of coordinating time and event order due to separate clocks and lack of shared memory. It reviews various synchronization algorithms, including Cristian's, Berkeley, Lamport's Logical Clocks, and Vector Clocks, each with their advantages and limitations. Practical applications such as totally ordered multicasting and mutual exclusion are also examined, emphasizing the need for correct ordering and concurrency detection in systems like chat applications.

Uploaded by

Alemayehu Kumsa
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
2 views3 pages

Distributed Systems Synchronization Assignment

The document discusses synchronization in distributed systems, highlighting the importance of coordinating time and event order due to separate clocks and lack of shared memory. It reviews various synchronization algorithms, including Cristian's, Berkeley, Lamport's Logical Clocks, and Vector Clocks, each with their advantages and limitations. Practical applications such as totally ordered multicasting and mutual exclusion are also examined, emphasizing the need for correct ordering and concurrency detection in systems like chat applications.

Uploaded by

Alemayehu Kumsa
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Distributed Systems – Synchronization Assignment

Part 1: Conceptual Understanding

1. Introduction to Synchronization

- Synchronization refers to coordinating time, event order, or shared resource access in


distributed systems. It is essential because distributed nodes have separate clocks and no
shared memory.

- Synchronization ensures correct coordination by maintaining consistent event ordering.

2. Issues in Synchronization

- No global clock and no shared memory cause communication delays and inconsistent local
clocks.

- Clock skew: difference in current time between clocks.

- Clock drift: difference in the rate clocks run. Both cause synchronization errors.

Part 2: Algorithms

Cristian’s Algorithm

- Client requests time from server.

- Server sends time.

- Client calculates delay and adjusts clock.

- Issues: network delay, server failure, trust in server.

Berkeley Algorithm

- Master polls clocks, averages time, sends adjustments.

- Differs from Cristian’s because it uses internal averaging.


- Advantages: no external server, fault tolerance. Limitations: master failure.

Lamport’s Logical Clocks

- Logical clocks track event order.

- Happened-before relation defines causality.

- Ensures consistency by incrementing timestamps and comparing message timestamps.

Vector Clocks

- Array of counters representing each process.

- Address Lamport’s limitation by detecting concurrency.

Part 3: Practical Scenarios

Totally Ordered Multicasting

- Prevents inconsistent updates across locations by enforcing identical message order.

Mutual Exclusion

- Critical section problem requires only one process to access shared resource.

- Centralized: simple but single point failure.

- Decentralized: voting but disagreement possible.

- Distributed: many messages required.

Part 4: Critical Analysis

Lamport’s failures occur when detecting concurrency; vector clocks fix this.

Vector clocks are best when full causality is required.


Real-time gaming requires Cristian’s/NTP because they use real physical time.

Part 5: Application

Chat system example using vector clocks ensures correct ordering and concurrency
detection.

You might also like