Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 1: Definition, Synchronous and Asynchronous Execution
Synchronous Execution: All processes proceed in a lockstep manner using a global clock.
Asynchronous Execution: Processes proceed independently with no global clock. Timing of events can vary.
Used in real-world distributed systems where delays and variations are normal.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 2: Model of Distributed Computations
Defines how processes and communication are represented. Models include process graphs, event ordering
(Lamport clocks), and channels.
Important for designing and analyzing distributed algorithms.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 3: Causality and Time in Distributed Systems
Causality: The relationship between events where one can influence another.
Lamport Clocks: Logical clocks used to order events in distributed systems.
Vector Clocks: Extended version to track causality more precisely.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 4: Graph Algorithms
Used to solve problems like shortest path, spanning trees, etc., in distributed systems.
Important for network topology, routing, and data dissemination.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 5: Message Ordering and Group Communication
Ensures consistent communication in distributed systems.
Includes FIFO, Causal, and Total ordering.
Group Communication ensures messages reach all members reliably and in order.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 6: Coordination Algorithms (T Book 5)
Includes mutual exclusion (Ricart-Agrawala, Token Ring), election (Bully, Ring), and barrier synchronization.
Ensures orderly access to resources and coordination among distributed processes.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 7: Coordination Algorithms (T Book 2)
Expands on coordination using quorum, timeouts, and middleware like ZooKeeper.
Deadlock detection and resource allocation strategies also covered.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 8: Consistency and Replication
Replication maintains data copies across nodes.
Consistency models include Strong, Sequential, Causal, and Eventual.
Protocols like quorum, primary-backup ensure data consistency and availability.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 9: Global State and Snapshot Recording Algorithms
Chandy-Lamport algorithm captures a consistent snapshot of the system.
Used for debugging, checkpointing, and detecting global properties.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 10: Self-Stabilizing Distributed System
Automatically recovers from arbitrary faults to a legal state.
Ensures convergence and closure.
Used in network recovery, sensor networks, etc.
Parallel and Distributed Computing (PDC) - 50 Marks Syllabus Notes
Topic 11: Fault-Tolerant Message-Passing Systems
Continues functioning despite process or message failures.
Uses retries, acknowledgements, consensus (Paxos, Raft), and failure detectors.
Critical for robust distributed systems like databases and cloud platforms.