Coordination in distributed systems

This document summarizes key concepts related to distributed mutual exclusion and distributed deadlock detection. It discusses classification of distributed mutual exclusion algorithms into token-based and non-token-based approaches. For distributed mutual exclusion, it describes Lamport's algorithm, Ricart-Agrawala algorithm, Maekawa's quorum-based algorithm, and Suzuki-Kasami's token-based broadcast algorithm. It also discusses requirements for mutual exclusion such as freedom from deadlock and starvation. For distributed deadlock detection, it mentions the system model and types of deadlocks as well as approaches for prevention, avoidance, detection, and resolution of deadlocks.

Diego Souza fala sobre sistemas distribuídos mostradando uma introdução sobre os conceitos básicos e algumas considerações práticas que podem afetar o nosso dia a dia. Assista esta palestra em https://www.eventials.com/locaweb/sistemas-distribuidos/

The reader/writer problem involves coordinating access to shared data by multiple reader and writer processes. There are two main approaches: (1) prioritizing readers, where readers can access the data simultaneously but writers must wait, risking writer starvation. This can be solved using semaphores. (2) Prioritizing writers, where new readers must wait if a writer is already accessing the data. This prevents starvation and can be implemented using monitors. The document then describes how to use semaphores to solve the reader/writer problem by prioritizing readers, with mutex, wrt, and readcount semaphores controlling access for readers and writers.

Gossip là một giao thức trao đổi thông tin phổ biến trong các hệ thống phân tán giúp cho các máy chủ duy trì trạng thái đồng nhất với nhau cũng như thực hiện các nhiệm vụ có chủ đích. Điểm mạnh của nó là khả năng phát tán thông tin ở tốc độ cao cũng như không hề có single point of failure. Trong bài talk này, Anh Nguyễn Anh Tú, thành viên của Grokking sẽ chia sẻ một số thông tin về giao thức Gossip cũng như điểm qua một vài ứng dụng thực tiễn của nó. - Về diễn giả: Anh Nguyễn Anh Tú hiện đang là Staff Software Engineer tại Axon Vietnam, đồng thời là thành viên của Grokking Vietnam.

Concurrency Control Techniques: Concurrency Control, Locking Techniques for Concurrency Control, Time Stamping Protocols for Concurrency Control, Validation Based Protocol, Multiple Granularity, Multi Version Schemes, Recovery with Concurrent Transaction,

This document discusses task migration in distributed systems. It defines task migration as the preemptive transfer of a partially executed task to another node. The document outlines the group members working on the task migration project and provides an introduction to terminology used. It then discusses reasons for task migration like load distribution and improving communication. The main issues discussed are state transfer costs, location transparency, migration mechanism structure, and performance impacts.

Replication in computing involves sharing information so as to ensure consistency between redundant resources, such as software or hardware components, to improve reliability, fault-tolerance, or accessibility.

Trafodion brings a completely distributed scalable transaction management implementation integrated into HBase. It does not suffer from the scale and performance limitations of other transaction managers on HBase. This presentation reviews the elegant architecture and how this architecture is leveraged to provide full ACID SQL transactional capabilities across multiple rows, tables, statements, and region servers. It discusses the life of a transaction from BEGIN WORK, to updates, to ABORT WORK, to COMMIT WORK, and then discusses recovery and high availability capabilities provided. An accompanying white paper goes into depth explaining this animated presentation in more detail. Given the increasing interest for transaction managers on Hadoop, or to provide transactional capabilities for NoSQL users when needed, the Trafodion community can certainly open up this Distributed Transaction Management support to be leveraged by implementations other than Trafodion.

In 1965, Dijkstra proposed a new and very significant technique for managing concurrent processes by using the value of a simple integer variable to synchronize the progress of interacting processes. This integer variable is called semaphore. So it is basically a synchronizing tool and is accessed only through two low standard atomic operations, wait and signal designated by P() and V() respectively. The classical definition of wait and signal are : Wait : decrement the value of its argument S as soon as it would become non-negative. Signal : increment the value of its argument, S as an individual operation.