My Account Log in

1 option

Middleware 2011 : ACM/IFIP/USENIX 12th International Middleware Conference, Lisbon, Portugal, December 12-16, 2011, Proceedings / edited by Fabio Kon, Anne-Marie Kermarrec.

ACM Digital Library Available online

View online
Format:
Book
Contributor:
Kon, Fabio, Editor.
Kermarrec, Anne-Marie, Editor.
Series:
Programming and Software Engineering, 2945-9168 ; 7049
Language:
English
Subjects (All):
Computer networks.
Application software.
Software engineering.
Information storage and retrieval systems.
Electronic data processing--Management.
Electronic data processing.
Computers and civilization.
Computer Communication Networks.
Computer and Information Systems Applications.
Software Engineering.
Information Storage and Retrieval.
IT Operations.
Computers and Society.
Local Subjects:
Computer Communication Networks.
Computer and Information Systems Applications.
Software Engineering.
Information Storage and Retrieval.
IT Operations.
Computers and Society.
Physical Description:
1 online resource (XIII, 514 p.)
Edition:
1st ed. 2011.
Place of Publication:
Berlin, Heidelberg : Springer Berlin Heidelberg : Imprint: Springer, 2011.
Language Note:
English
Summary:
This book constitutes the refereed proceedings of the ACM/IFIP/USENIX 12th International Middleware Conference, held in Lisbon, Portugal, in December 2011. The 22 revised full papers presented together with 2 industry papers and an invited paper were carefully reviewed and selected from 125 submissions. The papers are organized in topical sections on social networks, storage and performance management, green computing and resource management, notification and streaming, replication and caching, security and interoperability, and run-time (re)configuration and inspection.
Contents:
Intro
Title
Preface
Organization
Table of Contents
Invited Paper
Democratizing Transactional Programming
A Brief History of Transaction
The Inherent Appeal of Transactions
Preserving Sequentiality
Enabling Composition
The Inherent Limitations of Transactions
Lacking Expressiveness
Impact on Concurrency
Impact on Performance
Democratizing Transactions: The Challenge
Expressiveness and Simplicity
Sequentiality and Composition
Mixing Several Semantics
Combining Classic, Snapshot and Elastic Transactions
Concluding Remarks
References
Social Networks
Scaling Microblogging Services with Divergent Traffic Demands
Introduction
Background and Related Work
Measuring Availability at High Load
Design
System Architecture
Social Relations
Unicast Delivery for the Social Network
Gossip Dissemination for the Media Network
Support for Client Heterogeneity
Message Loss Detection and Security Issues
Experimental Evaluation
Experiment Settings
Server Cloud Performance
Cuckoo Peer Performance
Conclusion
Contrail: Enabling Decentralized Social Networks on Smartphones
Problem Statement
Design of Contrail
Sender-Side Filters
Cloud Relays
Reliability and Security in Contrail
The Contrail System
Contrail on the Phone
Contrail in the Cloud
Applications
The Location Notification Application
Potential Contrail Applications
Evaluation
End-to-End Latency
Contrail Scalability
Contrail Throughput
Energy Consumption and Filtering
Related Work
Confidant: Protecting OSN Data without Locking It Up
Overview
Architecture
Trust and Threat Model
Cryptographic State.
Objects and Access Policies
Name Servers
Storage Servers
Implementation
Client
Name Server
Storage Server
Untrusted Storage Server
Application Performance
Availability
Storage and Performance Management
Live Deduplication Storage of Virtual Machine Images in an Open-Source Cloud
LiveDFS Design
Primitives
Deduplication Design
Prefetching of Fingerprint Store
Journaling
LiveDFS Implementation and Deployment
Experiments
I/O Throughput
OpenStack Deployment
Conclusions and Future Work
Scalable Load Balancing in Cluster Storage Systems
Preliminaries
Algorithms for Load Balancing
Exact Solution: ILP Formulation
Approximation
Implementation of Ursa on Windows Azure
Evaluation on Simulator
Evaluation on Windows Azure
Conclusion and Future Work
Predico: A System for What-if Analysis in Complex Data Center Applications
Background and Problem Formulation
Modeling a Data Center Application
Queuing Theoretic Node-Level Models
Workload Models
Model Composition: From Node-Level to System-Level Models
Answering What-if Queries
On-the-Fly Model Construction Using the Influence Graph
Query Execution Using Change Propagation
Predico Implementation
Posing What-if Queries in Predico
Prototype Implementation
Experimental Setup
Accuracy of Node-Level Resource Usage and Latency Models
Accuracy of Node-Level Workload Models
Accuracy of System-Level Models with Increasing Composition Depth
Accuracy of System-Level Models with Varying Topology
Workload-Only What-If Analysis Case Study
Conclusions.
References
Green Computing and Resource Management
GreenWare: Greening Cloud-Scale Data Centers to Maximize the Use of Renewable Energy
GreenWare Architecture
Design Methodology of GreenWare
Problem Formulation
Response Time and Power Models
Wind Power Model
Solar Power Model
Problem Solution
Simulation Setup
Datacenter Parameters
Renewable Energy Availability
Real-World Workload Traces
Electricity Price Traces
Evaluation Results
Baselines
Impacts of the Monthly Cost Budget
Comparison with Baselines
Impacts of Pricing Policies of Renewable Energy
Resource Provisioning Framework for MapReduce Jobs with Performance Goals
MapReduce Background
Profiling MapReduce Jobs
Job Execution as a Function of Allocated Resources
Job Performance Invariants as a Job Profile
MapReduce Performance Model
General Theoretical Bounds
Bounds-Based Completion Time Estimates of a MapReduce Job
Scaling Factors
Impact of Failures on the Completion Time Bounds
SLO-Based Resource Provisioning
Performance Invariants
Performance Bounds of Job Completion Times
Prediction of Job Completion Time with Failures
Resource-Aware Adaptive Scheduling for MapReduce Clusters
MapReduce and Hadoop
Resource-Aware Adaptive Scheduler
Performance Model
Placement Algorithm and Optimization Objective
Task Scheduler
Job Profiles
Experimental Environment and Workload
Experiment 1: Execution with Relaxed Completion Time Goals
Experiment 2: Execution with Tight Completion Time Goals
Conclusions
References.
Notification and Streaming
A Content-Based Publish/Subscribe Matching Algorithm for 2D Spatial Objects
Background
Spatial Concepts and Spatial Modeling
The Siena Fast Forwarding Algorithm
Spatial Model
A New 4-Intersection Model
Algorithm
Boundary Discretization
Complexity
Indexing and Simplification
Spatial Index
Simplification
Worst-Case Performance and Scalability
Accuracy and Performance in Realistic Configurations
FAIDECS: Fair Decentralized Event Correlation
Events: composition and correlation.
Challenges for event correlation middleware.
Contributions.
FAIDECS Model
Predicate Grammar
Predicate Types and Evaluation
Properties
Total Order
Algorithms
Total Order Broadcast Black Box
FAIDECS Decentralized Ordered Merging
Metrics and Experimental Setup
Conjunctions
Disjunctions
AmbiStream: A Middleware for Multimedia Streaming on Heterogeneous Mobile Devices
Challenges for Mobile Interoperable Media Streaming
The Streaming Process
Streaming Protocol Heterogeneity
Media Container Adaptation
AmbiStream Architecture
Streaming Protocol Translation
Media Container Format Adaptation
Experimental Results
Collecting Mobile Device Performance Data
Translating to RTSP between Android Smartphones
Translating to HLS between Android and iOS Smartphones
Virtualizing Stream Processing
Exoengine Architecture
Layers
Data Processing Model
Implementation Model
Component Life Cycle Management
Stream Processing with the Exoengine Platform
Porting Existing Stream Processing Engines.
Extensibility
Platform Implementation
Component Implementation
Application Builders
Distributed Operation
Prototype
The Linear Road Benchmark
Experiment Setup
Porting MXQuery and Linear Road
Overhead of the Exoengine Architecture
Replacing an Slet at Runtime
Distributed Deployment
Developing with the Exoengine
Heterogeneity
Discussion and Outlook
Replication and Caching
Leader Election for Replicated Services Using Application Scores
Application Scores
Background: Replicated Coordination Services
Request Latency
Recovery Time
The POLE Algorithm
Failure Detection
Leader Selection Algorithm
Oracles
Experimental Setup and Methodology
Request Latency and Random Request Distributions
Request Latency and Uneven Request Distributions
Extensions
PolyCert: Polymorphic Self-optimizing Replication for In-Memory Transactional Grids
Motivations
The PolyCert Replication Protocol
Replication Protocol Selection Oracle
A Trigger-Based Middleware Cache for ORMs
Caching Abstractions
Database Triggers
Consistency Guarantees
Programmer Effort
Microbenchmarks
Social Networking Workload
Security and Interoperability
Deploy, Adjust and Readjust: Supporting Dynamic Reconfiguration of Policy Enforcement
Background and Example
The Reference Enforcement Model
Illustrative Example
Attribute Configuration
Attribute Retrieval.
Attribute Caching.
Notes:
Bibliographic Level Mode of Issuance: Monograph
Includes bibliographical references and index.
ISBN:
3-642-25821-2
OCLC:
769755543

The Penn Libraries is committed to describing library materials using current, accurate, and responsible language. If you discover outdated or inaccurate language, please fill out this feedback form to report it and suggest alternative language.

My Account

Shelf Request an item Bookmarks Fines and fees Settings

Guides

Using the Library Catalog Using Articles+ Library Account