HAUPTSEMINAR BETRIEBSSYSTEME (500070)

Save this PDF as:
 WORD  PNG  TXT  JPG

Size: px
Start display at page:

Download "HAUPTSEMINAR BETRIEBSSYSTEME (500070)"

Transcription

1 HAUPTSEMINAR BETRIEBSSYSTEME (500070)! OPERATING SYSTEMS SEMINAR! CONSENSUS AND CONSISTENCY IN MODERN DISTRIBUTED SYSTEMS Dr. Peter Tröger Group, TU Chemnitz, Winter Term 2014/15

2 COURSE ORGANIZATION Study regulations Student presentation (30 minutes) + discussion (15 minutes) Written report (8-15 pages) Individual work, no groups Scheduling Initial topic presentation + choice of topics today Block seminar with presentations in December 2014 (tbd) Deadline for written reports in January 2015 (tbd) Bi-weekly individual consultation, appointed by 2

3 HOW NOT TO FAIL Grading depends on presentation and report ( 50% / 50% ) Respect limits (presentation time and report length) Give a great oral presentation Focus on comprehensible story line Focus on important ideas, e.g. crucial steps in a proof Be prepared for the discussion part Write a great report Good language, proper document style, correct citing Use your own words, don t just copy facts Bi-weekly consultation must be driven by you 3

4 REPORT STRUCTURE Introduction Description of the application field General problem to be solved, history of it Related Work Other approaches / solutions similar to yours [Main part] All the details about your particular topic [Experiments] Things you tried out, measurement results, experiences Summary 4

5 Consensus and Consistency in Modern Distributed Systems

6 DISTRIBUTED SYSTEMS Abstract view Nodes ( processes ) exchanging messages, beside that decoupled Partial failures of nodes, impact on messaging and/or internal state Synchronous model Known upper limit on message transmission time If the message is missing, the other node is gone Asynchronous model Unknown upper limit on message transmission Fits to reality 6

7 DISTRIBUTED SYSTEMS Scalability Handle more clients or internal load by adding nodes Availability Uptime / Lifetime Get more uptime by adding redundant nodes Modern distributed systems need both things Google, Twitter, Amazon, [microservices.io] 7

8 DISTRIBUTED SYSTEM Problems in distributed systems Mutual exclusion for shared resource access Concurrent access to shared resource in distributed system Critical section problem - like in OS, but no shared memory Election of leader Choose a process to play a particular role Difference to mutual exclusion - leader must be known to everybody, and does not give away leadership explicitly after some time Ordered multicast Group of processes should receive copies of message sent to the group, sometimes with delivery and order guarantees 8

9 EXAMPLE: MUTUAL EXCLUSION Assumptions: Processes do not fail, reliable message delivery, at-most-once Ensure exclusive access in critical section No deadlocks and starvation, fairness, message passing First idea: Central server algorithm Central server grants permission by providing token -> lock service Message types: Request token, release token, grant token Requests are queued on the central server, clients are therefore blocked Example: File locking daemon lockd in NFS 9

10 EXAMPLE: NFS LOCKD RPC-based Network Lock Manager (NLM) service rpc.lockd daemon on both client and server side Application issues fctl() command for file locking, kernel sends request to local lock daemon (Kernel Lock Manager protocol) Forwarded to remote host, 5 asynchronous operations Test for lock (returns conflicting lock, or LCK_GRANTED) Claim a lock (LCK_GRANTED or LCK_DENIED / LCK_BLOCKED) Unlock a lock Cancel a blocked lock request (client is no longer interested) Grant a blocked lock (notification of server to client) 10

11 EXAMPLE: MUTUAL EXCLUSION Second idea: Ring-based algorithm Arrange agreement without additional node Unidirectional ring topology, unrelated to physical interconnection One token per protected resource, forwarded if no interest Server vs. ring approach: Performance (in number of messages) Consumed bandwidth per reservation Client delay for critical section entry and exit Synchronization delay between one process leaving and the next one entering the critical section 11

12 FAILURES? Lost messages Not tolerated by algorithms, must be solved by other means Crashed process Not tolerated by ring-based approaches Server can tolerate client crash if it neither holds or requested the token Workarounds by book-keeping (see NFS statd) Even if faulty process can be detected, the system state must be re-established Example: Determine current owner of the token in server-based approach 12

13 FAILURE DETECTORS Fact: Things break in distributed systems Generalized idea of failure detectors Node queries if a particular other node failed Should never suspect another node unless it failed (accuracy) Eventually permanently suspects a crashed node (completeness) Unreliable failure detector states unsuspected or suspected Example implementation: pair-wise heartbeat Too short timeout vs. too long timeout Reliable failure detector states unsuspected or failed Foundation for consensus protocols 13

14 CONSENSUS Fact: Things break in distributed systems Problem: Maintain a shared state, even with node failures Example: Distributed storage system What is our (transaction) log of recent activities? Shall we commit this value now? Who has this data item locked? Who is the master? Idea: All nodes propose a value and must agree on one of them Error-free nodes get agreement, regardless of what the others say 14

15 CONSENSUS Consensus: Demand for coordination in system error case Example for fault model [Christian]!!!!!! Any obvious solution? No, it s not two-phase commit 15

16 TWO-PHASE COMMIT Coordinator Commit? Sure! Commit? Sure! A B Coordinator Commit! Commit! A B 16

17 TWO GENERALS PROBLEM Credited to Jim Gray, developed by Akkoyunlu et al. Thought experiment for coordination with unreliable links Two armies with generals prepare attack on city, each on a hill Only communication path through a valley, occupied by defenders Agreement on attack exists, but not on time Acknowledgment of receipt is needed Proven that no algorithms can be designed to solve the problem Sending acknowledgment for the acknowledgment for the acknowledgment... (every message could be modified) Termination on number of steps makes the last message the relevant one But sender of last message still lacks trustworthy acknowledgment 17

18 CONSENSUS Vast body of theory work about the problem Several outstanding contributions (Real) distributed system = asynchronous messaging = problem FLP impossibility proof Achieving consensus in the real asynchronous world, anyway Leslie Lamport s Byzantine Agreement Leslie Lamport s Paxos Raft (Hint: This works only with additional restrictions) 18

19 CONSISTENCY Consensus assumption Non-faulty nodes send messages that may be delayed Nodes not sending messages are failed Real world: Partitioning Non-faulty nodes may unable to send messages Should still maintain consistency of internal state Description of options in the CAP theorem [blog.mongodb.org] 19

20 CAP THEOREM CA - Classical DBMS, partitions are critical C Consistent view for clients Availability for clients A CP - Decreased availability for some clients P Partition Tolerance AP - Eventual consistency for clients 20

21 EVENTUAL CONSISTENCY [cloud.google.com] Consistency from client perspective Nothing new, remember principles of domain name system (DNS) 21

22 EVENTUAL CONSISTENCY Different projects apply eventual consistency to deal with CAP Amazon s Dynamo: Distributed key-value store, HA focus Apache Cassandra: Distributed DBMS, configurable replication and consistency Yahoo PNUTS: Configurable consistency, geographic distribution consideration Apache Storm Riak Voldemort 22

23 SEMINAR TOPICS Consensus Failure detectors FLP impossibility Paxos in theory and practice Raft! Consistency CAP Theorem Eventual consistency Amazon Dynamo Apache Cassandra Yahoo PNUTS Apache Storm Riak Voldemort Literature starting points on course homepage! 23

CSE-E5430 Scalable Cloud Computing Lecture 6

CSE-E5430 Scalable Cloud Computing Lecture 6 CSE-E5430 Scalable Cloud Computing Lecture 6 Keijo Heljanko Department of Computer Science School of Science Aalto University keijo.heljanko@aalto.fi 26.10-2015 1/20 Hard Disk Read Errors Unrecoverable

More information

Dr Markus Hagenbuchner markus@uow.edu.au CSCI319. Distributed Systems

Dr Markus Hagenbuchner markus@uow.edu.au CSCI319. Distributed Systems Dr Markus Hagenbuchner markus@uow.edu.au CSCI319 Distributed Systems CSCI319 Chapter 8 Page: 1 of 61 Fault Tolerance Study objectives: Understand the role of fault tolerance in Distributed Systems. Know

More information

Big Data Management and NoSQL Databases

Big Data Management and NoSQL Databases NDBI040 Big Data Management and NoSQL Databases Lecture 4. Basic Principles Doc. RNDr. Irena Holubova, Ph.D. holubova@ksi.mff.cuni.cz http://www.ksi.mff.cuni.cz/~holubova/ndbi040/ NoSQL Overview Main objective:

More information

Distributed Systems. Tutorial 12 Cassandra

Distributed Systems. Tutorial 12 Cassandra Distributed Systems Tutorial 12 Cassandra written by Alex Libov Based on FOSDEM 2010 presentation winter semester, 2013-2014 Cassandra In Greek mythology, Cassandra had the power of prophecy and the curse

More information

Middleware and Distributed Systems. System Models. Dr. Martin v. Löwis. Freitag, 14. Oktober 11

Middleware and Distributed Systems. System Models. Dr. Martin v. Löwis. Freitag, 14. Oktober 11 Middleware and Distributed Systems System Models Dr. Martin v. Löwis System Models (Coulouris et al.) Architectural models of distributed systems placement of parts and relationships between them e.g.

More information

NoSQL Databases. Institute of Computer Science Databases and Information Systems (DBIS) DB 2, WS 2014/2015

NoSQL Databases. Institute of Computer Science Databases and Information Systems (DBIS) DB 2, WS 2014/2015 NoSQL Databases Institute of Computer Science Databases and Information Systems (DBIS) DB 2, WS 2014/2015 Database Landscape Source: H. Lim, Y. Han, and S. Babu, How to Fit when No One Size Fits., in CIDR,

More information

Physicians are fond of saying Treat the problem, not the symptom. The same is true for Information Technology.

Physicians are fond of saying Treat the problem, not the symptom. The same is true for Information Technology. Comprehensive Consulting Solutions, Inc. Business Savvy. IT Smar Troubleshooting Basics: A Practical Approach to Problem Solving t. White Paper Published: September 2005 Physicians are fond of saying Treat

More information

DCaaS: Data Consistency as a Service for Managing Data Uncertainty on the Clouds

DCaaS: Data Consistency as a Service for Managing Data Uncertainty on the Clouds DCaaS: Data Consistency as a Service for Managing Data Uncertainty on the Clouds Islam Elgedawy Computer Engineering Department, Middle East Technical University, Northern Cyprus Campus, Guzelyurt, Mersin

More information

Scalability. We can measure growth in almost any terms. But there are three particularly interesting things to look at:

Scalability. We can measure growth in almost any terms. But there are three particularly interesting things to look at: Scalability The ability of a system, network, or process, to handle a growing amount of work in a capable manner or its ability to be enlarged to accommodate that growth. We can measure growth in almost

More information

Process groups and message ordering

Process groups and message ordering Process groups and message ordering If processes belong to groups, certain algorithms can be used that depend on group properties membership create ( name ), kill ( name ) join ( name, process ), leave

More information

Amr El Abbadi. Computer Science, UC Santa Barbara amr@cs.ucsb.edu

Amr El Abbadi. Computer Science, UC Santa Barbara amr@cs.ucsb.edu Amr El Abbadi Computer Science, UC Santa Barbara amr@cs.ucsb.edu Collaborators: Divy Agrawal, Sudipto Das, Aaron Elmore, Hatem Mahmoud, Faisal Nawab, and Stacy Patterson. Client Site Client Site Client

More information

Cassandra A Decentralized, Structured Storage System

Cassandra A Decentralized, Structured Storage System Cassandra A Decentralized, Structured Storage System Avinash Lakshman and Prashant Malik Facebook Published: April 2010, Volume 44, Issue 2 Communications of the ACM http://dl.acm.org/citation.cfm?id=1773922

More information

Distributed Data Stores

Distributed Data Stores Distributed Data Stores 1 Distributed Persistent State MapReduce addresses distributed processing of aggregation-based queries Persistent state across a large number of machines? Distributed DBMS High

More information

Introduction to NOSQL

Introduction to NOSQL Introduction to NOSQL Université Paris-Est Marne la Vallée, LIGM UMR CNRS 8049, France January 31, 2014 Motivations NOSQL stands for Not Only SQL Motivations Exponential growth of data set size (161Eo

More information

High Throughput Computing on P2P Networks. Carlos Pérez Miguel carlos.perezm@ehu.es

High Throughput Computing on P2P Networks. Carlos Pérez Miguel carlos.perezm@ehu.es High Throughput Computing on P2P Networks Carlos Pérez Miguel carlos.perezm@ehu.es Overview High Throughput Computing Motivation All things distributed: Peer-to-peer Non structured overlays Structured

More information

Do Relational Databases Belong in the Cloud? Michael Stiefel www.reliablesoftware.com development@reliablesoftware.com

Do Relational Databases Belong in the Cloud? Michael Stiefel www.reliablesoftware.com development@reliablesoftware.com Do Relational Databases Belong in the Cloud? Michael Stiefel www.reliablesoftware.com development@reliablesoftware.com How do you model data in the cloud? Relational Model A query operation on a relation

More information

Objectives. Distributed Databases and Client/Server Architecture. Distributed Database. Data Fragmentation

Objectives. Distributed Databases and Client/Server Architecture. Distributed Database. Data Fragmentation Objectives Distributed Databases and Client/Server Architecture IT354 @ Peter Lo 2005 1 Understand the advantages and disadvantages of distributed databases Know the design issues involved in distributed

More information

The Fail-Heterogeneous Architectural Model

The Fail-Heterogeneous Architectural Model The Fail-Heterogeneous Architectural Model Marco Serafini & Neeraj Suri TU Darmstadt Neeraj Suri EU-NSF ICT March 2006 Dependable Embedded Systems & SW Group www.deeds.informatik.tu-darmstadt.de Fault

More information

The Cloud Trade Off IBM Haifa Research Storage Systems

The Cloud Trade Off IBM Haifa Research Storage Systems The Cloud Trade Off IBM Haifa Research Storage Systems 1 Fundamental Requirements form Cloud Storage Systems The Google File System first design consideration: component failures are the norm rather than

More information

Unit 2 Distributed Systems R.Yamini Dept. Of CA, SRM University Kattankulathur

Unit 2 Distributed Systems R.Yamini Dept. Of CA, SRM University Kattankulathur Unit 2 Distributed Systems R.Yamini Dept. Of CA, SRM University Kattankulathur 1 Introduction to Distributed Systems Why do we develop distributed systems? availability of powerful yet cheap microprocessors

More information

Introduction to Distributed Systems

Introduction to Distributed Systems Introduction to Distributed Systems CSE 380 Computer Operating Systems Instructor: Insup Lee University of Pennsylvania Fall 2003 Lecture Note: Distributed Systems Why do we develop distributed systems?

More information

Web Email DNS Peer-to-peer systems (file sharing, CDNs, cycle sharing)

Web Email DNS Peer-to-peer systems (file sharing, CDNs, cycle sharing) 1 1 Distributed Systems What are distributed systems? How would you characterize them? Components of the system are located at networked computers Cooperate to provide some service No shared memory Communication

More information

MASTER PROJECT. Resource Provisioning for NoSQL Datastores

MASTER PROJECT. Resource Provisioning for NoSQL Datastores Vrije Universiteit Amsterdam MASTER PROJECT - Parallel and Distributed Computer Systems - Resource Provisioning for NoSQL Datastores Scientific Adviser Dr. Guillaume Pierre Author Eng. Mihai-Dorin Istin

More information

Availability Digest. MySQL Clusters Go Active/Active. December 2006

Availability Digest. MySQL Clusters Go Active/Active. December 2006 the Availability Digest MySQL Clusters Go Active/Active December 2006 Introduction MySQL (www.mysql.com) is without a doubt the most popular open source database in use today. Developed by MySQL AB of

More information

Transactions and ACID in MongoDB

Transactions and ACID in MongoDB Transactions and ACID in MongoDB Kevin Swingler Contents Recap of ACID transactions in RDBMSs Transactions and ACID in MongoDB 1 Concurrency Databases are almost always accessed by multiple users concurrently

More information

A SURVEY OF POPULAR CLUSTERING TECHNOLOGIES

A SURVEY OF POPULAR CLUSTERING TECHNOLOGIES A SURVEY OF POPULAR CLUSTERING TECHNOLOGIES By: Edward Whalen Performance Tuning Corporation INTRODUCTION There are a number of clustering products available on the market today, and clustering has become

More information

Blockchain, Throughput, and Big Data Trent McConaghy

Blockchain, Throughput, and Big Data Trent McConaghy Blockchain, Throughput, and Big Data Trent McConaghy Bitcoin Startups Berlin Oct 28, 2014 Conclusion Outline Throughput numbers Big data Consensus algorithms ACID Blockchain Big data? Throughput numbers

More information

Brewer s Conjecture and the Feasibility of Consistent, Available, Partition-Tolerant Web Services

Brewer s Conjecture and the Feasibility of Consistent, Available, Partition-Tolerant Web Services Brewer s Conjecture and the Feasibility of Consistent, Available, Partition-Tolerant Web Services Seth Gilbert Nancy Lynch Abstract When designing distributed web services, there are three properties that

More information

Big Data Development CASSANDRA NoSQL Training - Workshop. March 13 to 17-2016 9 am to 5 pm HOTEL DUBAI GRAND DUBAI

Big Data Development CASSANDRA NoSQL Training - Workshop. March 13 to 17-2016 9 am to 5 pm HOTEL DUBAI GRAND DUBAI Big Data Development CASSANDRA NoSQL Training - Workshop March 13 to 17-2016 9 am to 5 pm HOTEL DUBAI GRAND DUBAI ISIDUS TECH TEAM FZE PO Box 121109 Dubai UAE, email training-coordinator@isidusnet M: +97150

More information

Big Data JAMES WARREN. Principles and best practices of NATHAN MARZ MANNING. scalable real-time data systems. Shelter Island

Big Data JAMES WARREN. Principles and best practices of NATHAN MARZ MANNING. scalable real-time data systems. Shelter Island Big Data Principles and best practices of scalable real-time data systems NATHAN MARZ JAMES WARREN II MANNING Shelter Island contents preface xiii acknowledgments xv about this book xviii ~1 Anew paradigm

More information

A Brief Analysis on Architecture and Reliability of Cloud Based Data Storage

A Brief Analysis on Architecture and Reliability of Cloud Based Data Storage Volume 2, No.4, July August 2013 International Journal of Information Systems and Computer Sciences ISSN 2319 7595 Tejaswini S L Jayanthy et al., Available International Online Journal at http://warse.org/pdfs/ijiscs03242013.pdf

More information

Data Management in the Cloud

Data Management in the Cloud Data Management in the Cloud Ryan Stern stern@cs.colostate.edu : Advanced Topics in Distributed Systems Department of Computer Science Colorado State University Outline Today Microsoft Cloud SQL Server

More information

RAMCloud: Scalable Datacenter Storage Entirely in DRAM. John Ousterhout Stanford University

RAMCloud: Scalable Datacenter Storage Entirely in DRAM. John Ousterhout Stanford University RAMCloud: Scalable Datacenter Storage Entirely in DRAM John Ousterhout Stanford University Introduction New research project at Stanford Create large-scale storage systems entirely in DRAM Interesting

More information

Synchronization in. Distributed Systems. Cooperation and Coordination in. Distributed Systems. Kinds of Synchronization.

Synchronization in. Distributed Systems. Cooperation and Coordination in. Distributed Systems. Kinds of Synchronization. Cooperation and Coordination in Distributed Systems Communication Mechanisms for the communication between processes Naming for searching communication partners Synchronization in Distributed Systems But...

More information

Design Patterns for Distributed Non-Relational Databases

Design Patterns for Distributed Non-Relational Databases Design Patterns for Distributed Non-Relational Databases aka Just Enough Distributed Systems To Be Dangerous (in 40 minutes) Todd Lipcon (@tlipcon) Cloudera June 11, 2009 Introduction Common Underlying

More information

Survey on Comparative Analysis of Database Replication Techniques

Survey on Comparative Analysis of Database Replication Techniques 72 Survey on Comparative Analysis of Database Replication Techniques Suchit Sapate, Student, Computer Science and Engineering, St. Vincent Pallotti College, Nagpur, India Minakshi Ramteke, Student, Computer

More information

7 Distributed Key-Value-Stores. Prof. Dr. -Ing. Wolfgang Lehner

7 Distributed Key-Value-Stores. Prof. Dr. -Ing. Wolfgang Lehner 7 Distributed Key-Value-Stores Prof. Dr. -Ing. Wolfgang Lehner > Distributed Key/Value Stores Observation Many applications do not need a query language Instead primary key access only Restriction of functionality

More information

Distributed Systems: Concepts and Design

Distributed Systems: Concepts and Design Distributed Systems: Concepts and Design Edition 3 By George Coulouris, Jean Dollimore and Tim Kindberg Addison-Wesley, Pearson Education 2001. Chapter 2 Exercise Solutions 2.1 Describe and illustrate

More information

Distributed Data Management

Distributed Data Management Introduction Distributed Data Management Involves the distribution of data and work among more than one machine in the network. Distributed computing is more broad than canonical client/server, in that

More information

Synchronization in Distributed Systems

Synchronization in Distributed Systems Synchronization in Distributed Systems Chapter 4: Time and Synchronisation Page 1 1 Cooperation and Coordination in Distributed Systems Communication Mechanisms for the communication between processes

More information

Lecture Data Warehouse Systems

Lecture Data Warehouse Systems Lecture Data Warehouse Systems Eva Zangerle SS 2013 PART C: Novel Approaches in DW NoSQL and MapReduce Stonebraker on Data Warehouses Star and snowflake schemas are a good idea in the DW world C-Stores

More information

Principles and characteristics of distributed systems and environments

Principles and characteristics of distributed systems and environments Principles and characteristics of distributed systems and environments Definition of a distributed system Distributed system is a collection of independent computers that appears to its users as a single

More information

Cloud data store services and NoSQL databases. Ricardo Vilaça Universidade do Minho Portugal

Cloud data store services and NoSQL databases. Ricardo Vilaça Universidade do Minho Portugal Cloud data store services and NoSQL databases Ricardo Vilaça Universidade do Minho Portugal Context Introduction Traditional RDBMS were not designed for massive scale. Storage of digital data has reached

More information

Cloud Computing with Microsoft Azure

Cloud Computing with Microsoft Azure Cloud Computing with Microsoft Azure Michael Stiefel www.reliablesoftware.com development@reliablesoftware.com http://www.reliablesoftware.com/dasblog/default.aspx Azure's Three Flavors Azure Operating

More information

DISTRIBUTED SYSTEMS [COMP9243] Lecture 9a: Cloud Computing WHAT IS CLOUD COMPUTING? 2

DISTRIBUTED SYSTEMS [COMP9243] Lecture 9a: Cloud Computing WHAT IS CLOUD COMPUTING? 2 DISTRIBUTED SYSTEMS [COMP9243] Lecture 9a: Cloud Computing Slide 1 Slide 3 A style of computing in which dynamically scalable and often virtualized resources are provided as a service over the Internet.

More information

HA / DR Jargon Buster High Availability / Disaster Recovery

HA / DR Jargon Buster High Availability / Disaster Recovery HA / DR Jargon Buster High Availability / Disaster Recovery Welcome to Maxava s Jargon Buster. Your quick reference guide to Maxava HA and industry technical terms related to High Availability and Disaster

More information

Stretching A Wolfpack Cluster Of Servers For Disaster Tolerance. Dick Wilkins Program Manager Hewlett-Packard Co. Redmond, WA dick_wilkins@hp.

Stretching A Wolfpack Cluster Of Servers For Disaster Tolerance. Dick Wilkins Program Manager Hewlett-Packard Co. Redmond, WA dick_wilkins@hp. Stretching A Wolfpack Cluster Of Servers For Disaster Tolerance Dick Wilkins Program Manager Hewlett-Packard Co. Redmond, WA dick_wilkins@hp.com Motivation WWW access has made many businesses 24 by 7 operations.

More information

Geo-Replication in Large-Scale Cloud Computing Applications

Geo-Replication in Large-Scale Cloud Computing Applications Geo-Replication in Large-Scale Cloud Computing Applications Sérgio Garrau Almeida sergio.garrau@ist.utl.pt Instituto Superior Técnico (Advisor: Professor Luís Rodrigues) Abstract. Cloud computing applications

More information

Although research on distributed database systems. Consistency Tradeoffs in Modern Distributed Database System Design COVER FEATURE

Although research on distributed database systems. Consistency Tradeoffs in Modern Distributed Database System Design COVER FEATURE COVER FEATURE Consistency Tradeoffs in Modern Distributed Database System Design Daniel J. Abadi, Yale University The CAP theorem s impact on modern distributed database system design is more limited than

More information

Facebook: Cassandra. Smruti R. Sarangi. Department of Computer Science Indian Institute of Technology New Delhi, India. Overview Design Evaluation

Facebook: Cassandra. Smruti R. Sarangi. Department of Computer Science Indian Institute of Technology New Delhi, India. Overview Design Evaluation Facebook: Cassandra Smruti R. Sarangi Department of Computer Science Indian Institute of Technology New Delhi, India Smruti R. Sarangi Leader Election 1/24 Outline 1 2 3 Smruti R. Sarangi Leader Election

More information

Structured Data Storage

Structured Data Storage Structured Data Storage Xgen Congress Short Course 2010 Adam Kraut BioTeam Inc. Independent Consulting Shop: Vendor/technology agnostic Staffed by: Scientists forced to learn High Performance IT to conduct

More information

Distributed Systems, Failures, and Consensus. Jeff Chase Duke University

Distributed Systems, Failures, and Consensus. Jeff Chase Duke University Distributed Systems, Failures, and Consensus Jeff Chase Duke University The Players Choose from a large set of interchangeable terms: Processes, threads, tasks, Processors, nodes, servers, clients, Actors,

More information

BB2798 How Playtech uses predictive analytics to prevent business outages

BB2798 How Playtech uses predictive analytics to prevent business outages BB2798 How Playtech uses predictive analytics to prevent business outages Eli Eyal, Playtech Udi Shagal, HP June 13, 2013 Copyright 2013 Hewlett-Packard Development Company, L.P. The information contained

More information

Final Exam. Route Computation: One reason why link state routing is preferable to distance vector style routing.

Final Exam. Route Computation: One reason why link state routing is preferable to distance vector style routing. UCSD CSE CS 123 Final Exam Computer Networks Directions: Write your name on the exam. Write something for every question. You will get some points if you attempt a solution but nothing for a blank sheet

More information

Cloud Computing mit mathematischen Anwendungen

Cloud Computing mit mathematischen Anwendungen Cloud Computing mit mathematischen Anwendungen Vorlesung SoSe 2009 Dr. Marcel Kunze Karlsruhe Institute of Technology (KIT) Steinbuch Centre for Computing (SCC) KIT the cooperation of Forschungszentrum

More information

Cloud Computing Is In Your Future

Cloud Computing Is In Your Future Cloud Computing Is In Your Future Michael Stiefel www.reliablesoftware.com development@reliablesoftware.com http://www.reliablesoftware.com/dasblog/default.aspx Cloud Computing is Utility Computing Illusion

More information

Practical Cassandra. Vitalii Tymchyshyn tivv00@gmail.com @tivv00

Practical Cassandra. Vitalii Tymchyshyn tivv00@gmail.com @tivv00 Practical Cassandra NoSQL key-value vs RDBMS why and when Cassandra architecture Cassandra data model Life without joins or HDD space is cheap today Hardware requirements & deployment hints Vitalii Tymchyshyn

More information

Hadoop and Map-Reduce. Swati Gore

Hadoop and Map-Reduce. Swati Gore Hadoop and Map-Reduce Swati Gore Contents Why Hadoop? Hadoop Overview Hadoop Architecture Working Description Fault Tolerance Limitations Why Map-Reduce not MPI Distributed sort Why Hadoop? Existing Data

More information

Fault Tolerance. Ken Birman

Fault Tolerance. Ken Birman Fault Tolerance Ken Birman 1 Too many seminal concepts Lorenzo Alvisi s Byzantine twin wants you to use 2f+1 replicas Bonjour! Process pairs, primary-backup 2PC and 3PC, Quorums Atomic Transactions State

More information

High Availability with Postgres Plus Advanced Server. An EnterpriseDB White Paper

High Availability with Postgres Plus Advanced Server. An EnterpriseDB White Paper High Availability with Postgres Plus Advanced Server An EnterpriseDB White Paper For DBAs, Database Architects & IT Directors December 2013 Table of Contents Introduction 3 Active/Passive Clustering 4

More information

Communication System Design Projects

Communication System Design Projects Communication System Design Projects PROFESSOR DEJAN KOSTIC PRESENTER: KIRILL BOGDANOV KTH-DB Geo Distributed Key Value Store DESIGN AND DEVELOP GEO DISTRIBUTED KEY VALUE STORE. DEPLOY AND TEST IT ON A

More information

Distributed systems Lecture 6: Elec3ons, consensus, and distributed transac3ons. Dr Robert N. M. Watson

Distributed systems Lecture 6: Elec3ons, consensus, and distributed transac3ons. Dr Robert N. M. Watson Distributed systems Lecture 6: Elec3ons, consensus, and distributed transac3ons Dr Robert N. M. Watson 1 Last 3me Saw how we can build ordered mul3cast Messages between processes in a group Need to dis3nguish

More information

<Insert Picture Here> Oracle NoSQL Database A Distributed Key-Value Store

<Insert Picture Here> Oracle NoSQL Database A Distributed Key-Value Store Oracle NoSQL Database A Distributed Key-Value Store Charles Lamb, Consulting MTS The following is intended to outline our general product direction. It is intended for information

More information

Introduction to Big Data Training

Introduction to Big Data Training Introduction to Big Data Training The quickest way to be introduce with NOSQL/BIG DATA offerings Learn and experience Big Data Solutions including Hadoop HDFS, Map Reduce, NoSQL DBs: Document Based DB

More information

From the previous lecture

From the previous lecture CS 640: Introduction to Computer Networks Aditya Akella Lecture 7 - IP: Addressing and Forwarding From the previous lecture We will cover spanning tree from the last lecture 2 Spanning Tree Bridges More

More information

Cloud Based Distributed Databases: The Future Ahead

Cloud Based Distributed Databases: The Future Ahead Cloud Based Distributed Databases: The Future Ahead Arpita Mathur Mridul Mathur Pallavi Upadhyay Abstract Fault tolerant systems are necessary to be there for distributed databases for data centers or

More information

Boxwood: Abstractions as the Foundation for Storage Infrastructure

Boxwood: Abstractions as the Foundation for Storage Infrastructure Boxwood: Abstractions as the Foundation for Storage Infrastructure John MacCormick, Nick Murphy, Marc Najork, Chandramohan Thekkath, Lidong Zhou,Microsoft Research Presented by: Robert Robinson, CS848

More information

Client/Server and Distributed Computing

Client/Server and Distributed Computing Adapted from:operating Systems: Internals and Design Principles, 6/E William Stallings CS571 Fall 2010 Client/Server and Distributed Computing Dave Bremer Otago Polytechnic, N.Z. 2008, Prentice Hall Traditional

More information

Cloud Computing at Google. Architecture

Cloud Computing at Google. Architecture Cloud Computing at Google Google File System Web Systems and Algorithms Google Chris Brooks Department of Computer Science University of San Francisco Google has developed a layered system to handle webscale

More information

High Availability Essentials

High Availability Essentials High Availability Essentials Introduction Ascent Capture s High Availability Support feature consists of a number of independent components that, when deployed in a highly available computer system, result

More information

Data Center Networking

Data Center Networking Data Center Networking B4 Jain, Sushant, et al. "B4: Experience with a globally deployed software defined WAN." ACM SIGCOMM Computer Communication Review. Vol. 43. No. 4. ACM, 2013. (Google) B4 A private

More information

Cheap Paxos. Leslie Lamport and Mike Massa. Appeared in The International Conference on Dependable Systems and Networks (DSN 2004 )

Cheap Paxos. Leslie Lamport and Mike Massa. Appeared in The International Conference on Dependable Systems and Networks (DSN 2004 ) Cheap Paxos Leslie Lamport and Mike Massa Appeared in The International Conference on Dependable Systems and Networks (DSN 2004 ) Cheap Paxos Leslie Lamport and Mike Massa Microsoft Abstract Asynchronous

More information

CS Lecture 3 Network Architecture

CS Lecture 3 Network Architecture CS 557 - Lecture 3 Network Architecture End to End Arguments in System Design Saltzer, Reed, Clark, 1984 Design Philosophy of the DARPA Internet Protocols Clark 1988 Spring 2013 Architecture Dictionary

More information

David M. Kroenke and David J. Auer Database Processing:

David M. Kroenke and David J. Auer Database Processing: David M. Kroenke and David J. Auer Database Processing: Fundamentals, Design, and Implementation Chapter Nine: Managing Multiuser Databases 9-1 Chapter Objectives To understand the need for, and importance

More information

CSE-E5430 Scalable Cloud Computing Lecture 11

CSE-E5430 Scalable Cloud Computing Lecture 11 CSE-E5430 Scalable Cloud Computing Lecture 11 Keijo Heljanko Department of Computer Science School of Science Aalto University keijo.heljanko@aalto.fi 30.11-2015 1/24 Distributed Coordination Systems Consensus

More information

On- Prem MongoDB- as- a- Service Powered by the CumuLogic DBaaS Platform

On- Prem MongoDB- as- a- Service Powered by the CumuLogic DBaaS Platform On- Prem MongoDB- as- a- Service Powered by the CumuLogic DBaaS Platform Page 1 of 16 Table of Contents Table of Contents... 2 Introduction... 3 NoSQL Databases... 3 CumuLogic NoSQL Database Service...

More information

Apache Hadoop. Alexandru Costan

Apache Hadoop. Alexandru Costan 1 Apache Hadoop Alexandru Costan Big Data Landscape No one-size-fits-all solution: SQL, NoSQL, MapReduce, No standard, except Hadoop 2 Outline What is Hadoop? Who uses it? Architecture HDFS MapReduce Open

More information

Name: 1. CS372H: Spring 2009 Final Exam

Name: 1. CS372H: Spring 2009 Final Exam Name: 1 Instructions CS372H: Spring 2009 Final Exam This exam is closed book and notes with one exception: you may bring and refer to a 1-sided 8.5x11- inch piece of paper printed with a 10-point or larger

More information

TranScend. Next Level Payment Processing. Product Overview

TranScend. Next Level Payment Processing. Product Overview TranScend Next Level Payment Processing Product Overview Product Functions & Features TranScend is the newest, most powerful, and most flexible electronics payment system from INTRIX Technology, Inc. It

More information

ZooKeeper. Table of contents

ZooKeeper. Table of contents by Table of contents 1 ZooKeeper: A Distributed Coordination Service for Distributed Applications... 2 1.1 Design Goals...2 1.2 Data model and the hierarchical namespace...3 1.3 Nodes and ephemeral nodes...

More information

Data Consistency on Private Cloud Storage System

Data Consistency on Private Cloud Storage System Volume, Issue, May-June 202 ISS 2278-6856 Data Consistency on Private Cloud Storage System Yin yein Aye University of Computer Studies,Yangon yinnyeinaye.ptn@email.com Abstract: Cloud computing paradigm

More information

Evaluation of NoSQL databases for large-scale decentralized microblogging

Evaluation of NoSQL databases for large-scale decentralized microblogging Evaluation of NoSQL databases for large-scale decentralized microblogging Cassandra & Couchbase Alexandre Fonseca, Anh Thu Vu, Peter Grman Decentralized Systems - 2nd semester 2012/2013 Universitat Politècnica

More information

WSO2 Message Broker. Scalable persistent Messaging System

WSO2 Message Broker. Scalable persistent Messaging System WSO2 Message Broker Scalable persistent Messaging System Outline Messaging Scalable Messaging Distributed Message Brokers WSO2 MB Architecture o Distributed Pub/sub architecture o Distributed Queues architecture

More information

Cloud Computing. Chapter 1 Introducing Cloud Computing

Cloud Computing. Chapter 1 Introducing Cloud Computing Cloud Computing Chapter 1 Introducing Cloud Computing Learning Objectives Understand the abstract nature of cloud computing. Describe evolutionary factors of computing that led to the cloud. Describe virtualization

More information

Database Replication with MySQL and PostgreSQL

Database Replication with MySQL and PostgreSQL Database Replication with MySQL and PostgreSQL Fabian Mauchle Software and Systems University of Applied Sciences Rapperswil, Switzerland www.hsr.ch/mse Abstract Databases are used very often in business

More information

VMware Virtual SAN Remote Office / Branch Office Deployment

VMware Virtual SAN Remote Office / Branch Office Deployment SOLUTION OVERVIEW VMware Virtual SAN Radically Simple Storage for Remote and Branch Offices VMware Virtual SAN is VMware s radically simple, enterprise-class, software-defined storage solution for Hyper-Converged

More information

Databases : Lecture 11 : Beyond ACID/Relational databases Timothy G. Griffin Lent Term 2014. Apologies to Martin Fowler ( NoSQL Distilled )

Databases : Lecture 11 : Beyond ACID/Relational databases Timothy G. Griffin Lent Term 2014. Apologies to Martin Fowler ( NoSQL Distilled ) Databases : Lecture 11 : Beyond ACID/Relational databases Timothy G. Griffin Lent Term 2014 Rise of Web and cluster-based computing NoSQL Movement Relationships vs. Aggregates Key-value store XML or JSON

More information

Cluster Computing. ! Fault tolerance. ! Stateless. ! Throughput. ! Stateful. ! Response time. Architectures. Stateless vs. Stateful.

Cluster Computing. ! Fault tolerance. ! Stateless. ! Throughput. ! Stateful. ! Response time. Architectures. Stateless vs. Stateful. Architectures Cluster Computing Job Parallelism Request Parallelism 2 2010 VMware Inc. All rights reserved Replication Stateless vs. Stateful! Fault tolerance High availability despite failures If one

More information

NoSQL Data Base Basics

NoSQL Data Base Basics NoSQL Data Base Basics Course Notes in Transparency Format Cloud Computing MIRI (CLC-MIRI) UPC Master in Innovation & Research in Informatics Spring- 2013 Jordi Torres, UPC - BSC www.jorditorres.eu HDFS

More information

Chapter 8 Multiple Processor Systems. 8.1 Multiprocessors 8.2 Multicomputers 8.3 Distributed systems

Chapter 8 Multiple Processor Systems. 8.1 Multiprocessors 8.2 Multicomputers 8.3 Distributed systems Chapter 8 Multiple Processor Systems 8.1 Multiprocessors 8.2 Multicomputers 8.3 Distributed systems Multiprocessor Systems Continuous need for faster computers shared memory model message passing multiprocessor

More information

Multiprocessor Systems. Chapter 8 Multiple Processor Systems. Multiprocessors. Multiprocessor Hardware (1)

Multiprocessor Systems. Chapter 8 Multiple Processor Systems. Multiprocessors. Multiprocessor Hardware (1) Chapter 8 Multiple Processor Systems Multiprocessor Systems 8.1 Multiprocessors 8.2 Multicomputers 8.3 Distributed systems Continuous need for faster computers shared memory model message passing multiprocessor

More information

Distribution transparency. Degree of transparency. Openness of distributed systems

Distribution transparency. Degree of transparency. Openness of distributed systems Distributed Systems Principles and Paradigms Maarten van Steen VU Amsterdam, Dept. Computer Science steen@cs.vu.nl Chapter 01: Version: August 27, 2012 1 / 28 Distributed System: Definition A distributed

More information

Can the Elephants Handle the NoSQL Onslaught?

Can the Elephants Handle the NoSQL Onslaught? Can the Elephants Handle the NoSQL Onslaught? Avrilia Floratou, Nikhil Teletia David J. DeWitt, Jignesh M. Patel, Donghui Zhang University of Wisconsin-Madison Microsoft Jim Gray Systems Lab Presented

More information

Condor for the Grid. 3) http://www.cs.wisc.edu/condor/

Condor for the Grid. 3) http://www.cs.wisc.edu/condor/ Condor for the Grid 1) Condor and the Grid. Douglas Thain, Todd Tannenbaum, and Miron Livny. In Grid Computing: Making The Global Infrastructure a Reality, Fran Berman, Anthony J.G. Hey, Geoffrey Fox,

More information

Rakam: Distributed Analytics API

Rakam: Distributed Analytics API Rakam: Distributed Analytics API Burak Emre Kabakcı May 30, 2014 Abstract Today, most of the big data applications needs to compute data in real-time since the Internet develops quite fast and the users

More information

Application-specific databases have always been with us...

Application-specific databases have always been with us... Databases : Lecture 11 : Beyond ACID/Relational databases Timothy G. Griffin Lent Term 2014 Rise of Web and cluster-based computing NoSQL Movement Relationships vs. Aggregates Key-value store XML or JSON

More information

NoSQL. Thomas Neumann 1 / 22

NoSQL. Thomas Neumann 1 / 22 NoSQL Thomas Neumann 1 / 22 What are NoSQL databases? hard to say more a theme than a well defined thing Usually some or all of the following: no SQL interface no relational model / no schema no joins,

More information

An Open Source NoSQL solution for Internet Access Logs Analysis

An Open Source NoSQL solution for Internet Access Logs Analysis An Open Source NoSQL solution for Internet Access Logs Analysis A practical case of why, what and how to use a NoSQL Database Management System instead of a relational one José Manuel Ciges Regueiro

More information

Overview of Luna High Availability and Load Balancing

Overview of Luna High Availability and Load Balancing SafeNet HSM TECHNICAL NOTE Overview of Luna High Availability and Load Balancing Contents Introduction... 2 Overview... 2 High Availability... 3 Load Balancing... 4 Failover... 5 Recovery... 5 Standby

More information

Preparing Your Data For Cloud

Preparing Your Data For Cloud Preparing Your Data For Cloud Narinder Kumar Inphina Technologies 1 Agenda Relational DBMS's : Pros & Cons Non-Relational DBMS's : Pros & Cons Types of Non-Relational DBMS's Current Market State Applicability

More information