Data Sharing in OSD Environment Dingshan He September 30, 2002.

Slides:



Advertisements
Similar presentations
Distributed Systems Major Design Issues Presented by: Christopher Hector CS8320 – Advanced Operating Systems Spring 2007 – Section 2.6 Presentation Dr.
Advertisements

Replication. Topics r Why Replication? r System Model r Consistency Models r One approach to consistency management and dealing with failures.
MADFS: The Mobile Agent- based Distributed Network File system Presented by : Hailong Hou Instructor: Yanqing Zhang.
Distributed Databases John Ortiz. Lecture 24Distributed Databases2  Distributed Database (DDB) is a collection of interrelated databases interconnected.
DISTRIBUTED SYSTEMS II REPLICATION CNT. II Prof Philippas Tsigas Distributed Computing and Systems Research Group.
Cooperative Caching of Dynamic Content on a Distributed Web Server Vegard Holmedahl, Ben Smith, Tao Yang Speaker: SeungLak Choi, DB Lab., CS Dept.
Distributed Systems Fall 2010 Replication Fall 20105DV0203 Outline Group communication Fault-tolerant services –Passive and active replication Highly.
Database Replication techniques: a Three Parameter Classification Authors : Database Replication techniques: a Three Parameter Classification Authors :
Computer Science Lecture 14, page 1 CS677: Distributed OS Consistency and Replication Today: –Introduction –Consistency models Data-centric consistency.
On Reducing Communication Cost for Distributed Query Monitoring Systems. Fuyu Liu, Kien A. Hua, Fei Xie MDM 2008 Alex Papadimitriou.
2/23/2004 Load Balancing February 23, /23/2004 Assignments Work on Registrar Assignment.
Distributed Database Management Systems
Overview Distributed vs. decentralized Why distributed databases
OSD Metadata Management
Distributed Systems Fall 2009 Replication Fall 20095DV0203 Outline Group communication Fault-tolerant services –Passive and active replication Highly.
Object Naming & Content based Object Search 2/3/2003.
Concurrency Control & Caching Consistency Issues and Survey Dingshan He November 18, 2002.
Definition of terms Definition of terms Explain business conditions driving distributed databases Explain business conditions driving distributed databases.
University of Pennsylvania 11/21/00CSE 3801 Distributed File Systems CSE 380 Lecture Note 14 Insup Lee.
Client-Server Computing in Mobile Environments
Chapter 5 Roles and features. objectives Performing management tasks using the Server Manager console Understanding the Windows Server 2008 roles Understanding.
Databases with Scalable capabilities Presented by Mike Trischetta.
What is Architecture  Architecture is a subjective thing, a shared understanding of a system’s design by the expert developers on a project  In the.
Database Design – Lecture 16
Distributed Systems. Interprocess Communication (IPC) Processes are either independent or cooperating – Threads provide a gray area – Cooperating processes.
Consistency And Replication
Distributed File Systems
1 Adapted from Pearson Prentice Hall Adapted form James A. Senn’s Information Technology, 3 rd Edition Chapter 7 Enterprise Databases and Data Warehouses.
SAMANVITHA RAMAYANAM 18 TH FEBRUARY 2010 CPE 691 LAYERED APPLICATION.
Copyright © George Coulouris, Jean Dollimore, Tim Kindberg This material is made available for private study and for direct.
Week 5 Lecture Distributed Database Management Systems Samuel ConnSamuel Conn, Asst Professor Suggestions for using the Lecture Slides.
Distributed File Systems Overview  A file system is an abstract data type – an abstraction of a storage device.  A distributed file system is available.
CEPH: A SCALABLE, HIGH-PERFORMANCE DISTRIBUTED FILE SYSTEM S. A. Weil, S. A. Brandt, E. L. Miller D. D. E. Long, C. Maltzahn U. C. Santa Cruz OSDI 2006.
Data in the Cloud – I Parallel Databases The Google File System Parallel File Systems.
Module 7: Resolving NetBIOS Names by Using Windows Internet Name Service (WINS)
Introduction to DFS. Distributed File Systems A file system whose clients, servers and storage devices are dispersed among the machines of a distributed.
Distributed Computing Systems CSCI 4780/6780. Geographical Scalability Challenges Synchronous communication –Waiting for a reply does not scale well!!
Data Sharing. Data Sharing in a Sysplex Connecting a large number of systems together brings with it special considerations, such as how the large number.
Architecture Models. Readings r Coulouris, Dollimore and Kindberg Distributed Systems: Concepts and Design Edn. 3 m Note: All figures from this book.
Computer Science Lecture 13, page 1 CS677: Distributed OS Last Class: Canonical Problems Distributed synchronization and mutual exclusion Distributed Transactions.
Distributed OSes Continued Andy Wang COP 5611 Advanced Operating Systems.
Copyright © George Coulouris, Jean Dollimore, Tim Kindberg This material is made available for private study and for direct.
Caching Consistency and Concurrency Control Contact: Dingshan He
MQTT QoS2 Considerations Konstantin Dotchkoff. Challenges associated with implementing QoS 2 in large scale distributed systems Replication of QoS 2 messages.
GLOBAL EDGE SOFTWERE LTD1 R EMOTE F ILE S HARING - Ardhanareesh Aradhyamath.
Replication (1). Topics r Why Replication? r System Model r Consistency Models r One approach to consistency management and dealing with failures.
Memory Coherence in Shared Virtual Memory System ACM Transactions on Computer Science(TOCS), 1989 KAI LI Princeton University PAUL HUDAK Yale University.
Chapter 7: Consistency & Replication IV - REPLICATION MANAGEMENT By Jyothsna Natarajan Instructor: Prof. Yanqing Zhang Course: Advanced Operating Systems.
BACS 485 Multi-User Database Processing. Lecture Objectives Learn the difference between single and multi-user database processing and understand the.
1 Information Retrieval and Use De-normalisation and Distributed database systems Geoff Leese September 2008, revised October 2009.
Distributed Computing Systems Replication Dr. Sunny Jeong. Mr. Colin Zhang With Thanks to Prof. G. Coulouris,
Key Terms Windows 2008 Network Infrastructure Confiuguration Lesson 6
Replication.
CSI 400/500 Operating Systems Spring 2009
Chapter 7: Consistency & Replication IV - REPLICATION MANAGEMENT -Sumanth Kandagatla Instructor: Prof. Yanqing Zhang Advanced Operating Systems (CSC 8320)
CSE 451: Operating Systems Winter Module 22 Distributed File Systems
CSE 4340/5349 Mobile Systems Engineering
Distributed File Systems
Distributed File Systems
Outline Announcements Lab2 Distributed File Systems 1/17/2019 COP5611.
CSE 451: Operating Systems Spring Module 21 Distributed File Systems
SAMANVITHA RAMAYANAM 18TH FEBRUARY 2010 CPE 691
Distributed File Systems
CSE 451: Operating Systems Winter Module 22 Distributed File Systems
CS510 - Portland State University
DISTRIBUTED SYSTEMS Principles and Paradigms Second Edition ANDREW S
Outline Review of Quiz #1 Distributed File Systems 4/20/2019 COP5611.
Overview Multimedia: The Role of WINS in the Network Infrastructure
Distributed File Systems
Distributed File Systems
Presentation transcript:

Data Sharing in OSD Environment Dingshan He September 30, 2002.

Motivation & Issues OSD claims improved device & data sharing across platforms How does OSD facilitate data sharing? How to define data consistency semantics in the OSD environment? How to achieve data sharing following the defined data consistency semantics?

OSD Environment Network OSD Client replications of the same object Manager object access OPEN/CLOSE READ/WRITE MANAGEMENT

Region Concept There is a Region Manager represents the region RM’s communicate in the P2P manner OSD Client OSD Manager Region Network region boundaryP2P communication

Local Copy vs. Remote Copy From a single client’s point of view –Local copy: copy within the same region –Remote copy: copy within another region However, a client is unaware of whether the copy is local or remote A client always sends OPEN request to its regional manager, using a unique object id as parameter. The regional manager figures out where to find a nearby copy (either local or remote)

Primary/Secondary Copies Differentiate the importance of copies of the same object Facilitate concurrency control Open Question: –Is the role of primary copy static or dynamic?

Concurrency Control Multiple clients may simultaneously READ/WRITE the same object on a single OSD device, or multiple OSD devices Design Choices –Locking –Resynchronization

Locking Locking for one replication of an object could be performed at the regional manager of the OSD containing this replication Locking for different replications (stored on different OSD devices) of the same object –Either centralized by using a single manager –Or distributed by exchange information among managers

Locking Pros : –Applicable to any objects –Strong consistency Cons: –Slow response due to locking overhead –Potential deadlock –Increasing network traffic –Low availability with weak connectivity

Resynchronization Dealing with multiple clients simultaneously READ/WRITE different replications (on different OSD devices) of the same object Requirement: The resynchronization behavior of the object should be well defined and deterministic

Resynchronization Pros: –Quick response –Less network traffic –Higher availability –Suitable for weak connectivity environment Cons: –Not necessarily applicable to all kinds of objects –Weak consistency

Comments We are dealing with both strong connectivity and weak connectivity (such as mobile computing) environment Resynchronization is necessary We still want to achieve as much consistency as possible, so we can’t only rely on resynchronization

Conclusion Resynchronization and server locking should be both used Server locking is used to achieve as much consistency as possible Resynchronization is used to handle network partition condition, where complete consistency is just impossible, unless response time could be infinitely sacrificed

Levels of consistency Strong consistency: using server locking whenever object is accessed Weak consistency: using resynchronization Medium consistency: using both server locking and resynchronization

Object Support Object should have a way to indicate its desired consistency level For resynchronization, object should have its particular synchronization method, which will deterministically synchronize inconsistent replications of the same object

Caching Necessary!! Possible caching locations –Client –Manager Caching (temporary)  Replication (relatively permanent)

Caching Consistency Strong –Using revalidation –High latency Weak –Works like web caching –Potential inconsistency

Caching Granularity Big object Partial access Object could provide support by specifying the way to decompose itself –Example: a relational table object could decompose itself by rows.

Summary Object access in OSD environment Region concept Locking Resynchronization Caching