Automatic Load Balancing of ProActive Applications... with Fault Tolerance Javier Bustos, José Piquer, Universidad de Chile, Santiago, Chile Sebastien.

Slides:



Advertisements
Similar presentations
European Research Network on Foundations, Software Infrastructures and Applications for large scale distributed, GRID and Peer-to-Peer Technologies Experiences.
Advertisements

Christian Delbe1 Christian Delbé OASIS Team INRIA -- CNRS - I3S -- Univ. of Nice Sophia-Antipolis November Automatic Fault Tolerance in ProActive.
Scalable Content-Addressable Network Lintao Liu
Serverless Network File Systems. Network File Systems Allow sharing among independent file systems in a transparent manner Mounting a remote directory.
Topics in Grid Computing Orientation Sathish Vadhiyar.
Windows Server ® Virtualization Infrastructure Planning and Design Published: November 2007 Updated: July 2010.
Fabrice Huet Mobility of Active Objects Fabrice Huet INRIA-I3S-CNRS.
Mobile Communications in a Mobile Agent Based Overlay System Ching-Feng Li.
Web Caching Schemes1 A Survey of Web Caching Schemes for the Internet Jia Wang.
2/25/2004 The Google Cluster Architecture February 25, 2004.
City University London
490dp Synchronous vs. Asynchronous Invocation Robert Grimm.
© nCode 2000 Title of Presentation goes here - go to Master Slide to edit - Slide 1 Reliable Communication for Highly Mobile Agents ECE 7995: Term Paper.
2/18/2004 Challenges in Building Internet Services February 18, 2004.
Lee Center Workshop, May 19, 2006 Distributed Objects System with Support for Sequential Consistency.
Distributed Object System. Project Goals Develop a distributed system for performing time-consuming calculations. Load Balancing support. Fault Tolerance.
NetSolve Henri Casanova and Jack Dongarra University of Tennessee and Oak Ridge National Laboratory
VMware Update 2009 Daniel Griggs Solutions Architect, Virtualization Servers & Storage Solutions Practice Dayton OH.
Introspective Replica Management Yan Chen, Hakim Weatherspoon, and Dennis Geels Our project developed and evaluated a replica management algorithm suitable.
Technical Question Technical Question
A Grid-enabled Branch and Bound Algorithm for Solving Challenging Combinatorial Optimization Problems Authors: M. Mezmaz, N. Melab and E-G. Talbi Presented.
CS 443 Advanced OS Fabián E. Bustamante, Spring 2005 Porcupine: A Highly Available Cluster- based Mail Service Y. Saito, B. Bershad, H. Levy U. Washington.
Algorithms for Self-Organization and Adaptive Service Placement in Dynamic Distributed Systems Artur Andrzejak, Sven Graupner,Vadim Kotov, Holger Trinks.
Case Study - GFS.
Common Services in a network Server : provide services Type of Services (= type of servers) –file servers –print servers –application servers –domain servers.
ATIF MEHMOOD MALIK KASHIF SIDDIQUE Improving dependability of Cloud Computing with Fault Tolerance and High Availability.
Study of the Relationship between Peer to Peer Systems and IP Multicasting From IEEE Communication Magazine January 2003 學號 :M 姓名 : 邱 秀 純.
Load distribution in distributed systems
Windows Server ® Virtualization Infrastructure Planning and Design Published: November 2007 Updated: January 2012.
1 1 Hybrid Cloud Solutions (Private with Public Burst) Accelerate and Orchestrate Enterprise Applications.
MobSched: An Optimizable Scheduler for Mobile Cloud Computing S. SindiaS. GaoB. Black A.LimV. D. AgrawalP. Agrawal Auburn University, Auburn, AL 45 th.
PicsouGrid Viet-Dung DOAN. Agenda Motivation PicsouGrid’s architecture –Pricing scenarios PicsouGrid’s properties –Load balancing –Fault tolerance Perspectives.
Self Adaptivity in Grid Computing Reporter : Po - Jen Lo Sathish S. Vadhiyar and Jack J. Dongarra.
Module 12: Designing High Availability in Windows Server ® 2008.
MIT Consistent Hashing: Load Balancing in a Changing World David Karger, Eric Lehman, Tom Leighton, Matt Levine, Daniel Lewin, Rina Panigrahy.
Software Architecture Framework for Ubiquitous Computing Divya ChanneGowda Athrey Joshi.
Tsunami: Maintaining High Bandwidth Under Dynamic Network Conditions Dejan Kostić, Ryan Braud, Charles Killian, Eric Vandekieft, James W. Anderson, Alex.
DNS Zones. DNS records kept in zones DNS server is authoritative for a domain if it hosts the zone for that domain Sub-domains can be kept in same zone.
Frontiers in Massive Data Analysis Chapter 3.  Difficult to include data from multiple sources  Each organization develops a unique way of representing.
June 6, 2007TeraGrid '071 Clustering the Reliable File Transfer Service Jim Basney and Patrick Duda NCSA, University of Illinois This material is based.
Fast Searching in Peer-to-Peer Networks Self-Organizing Parallel Search Clusters Rocky Dunlap.
Fault Tolerant Extensions to Charm++ and AMPI presented by Sayantan Chakravorty Chao Huang, Celso Mendes, Gengbin Zheng, Lixia Shi.
Problem Wensheng Zhang, Dr. Guohong Cao, and Dr. Tom La Porta Example: Battlefield Surveillance Challenges Small Sensing Range Limitations in sensor nodes.
Windows Azure Virtual Machines Anton Boyko. A Continuous Offering From Private to Public Cloud.
 Cachet Technologies 1998 Cachet Technologies Technology Overview February 1998.
Chap 7: Consistency and Replication
+ Clusters Alternative to SMP as an approach to providing high performance and high availability Particularly attractive for server applications Defined.
NETWORK LOAD BALANCING (NLB) Microsoft Windows Server 2003 By Mohammad Alsawwaf ITEC452 Supervised By: Dr. Lee RADFORD UNIVERSITY.
1 Fault-Tolerant Mechanism for Hierarchical Branch and Bound Algorithm Université A/Mira de Béjaïa CEntre de Recherche sur l’Information Scientifique et.
Acknowledgement: These slides are adapted from slides provided in Thißen & Spaniol's course Distributed Systems and Middleware, RWTH Aachen Processes Distributed.
1 OASIS Team, INRIA Sophia-Antipolis/I3S CNRS, Univ. Nice Christian Delbé Data Grid Explorer 15/09/03 Large Scale Emulation Mobility in ProActive.
Fault Tolerance in Charm++ Gengbin Zheng 10/11/2005 Parallel Programming Lab University of Illinois at Urbana- Champaign.
Choosing an AS in a NutShell J.MOLIERE Who am I ? ► Independant author/consultant  Cahiers du programmeur Java – tome 2 – Eyrolles 2003  Cahiers.
By Ronnie Julio Mohammad Alsawwaf.  Using more than two computer systems that are linked together  Handles a larger/more variable workload  Provides.
The Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Presenter: Chao-Han Tsai (Some slides adapted from the Google’s series lectures)
IMPROVEMENT OF COMPUTATIONAL ABILITIES IN COMPUTING ENVIRONMENTS WITH VIRTUALIZATION TECHNOLOGIES Abstract We illustrates the ways to improve abilities.
Seminar On Rain Technology
System Models Advanced Operating Systems Nael Abu-halaweh.
Presented by Deepak Varghese Reg No: Introduction Application S/W for server load balancing Many client requests make server congestion Distribute.
TECHNICAL SEMINAR On. introduction  Cloud support for real time system is really important because, today we found a lot of real time systems around.
Scaling Network Load Balancing Clusters
Managing Multi-User Databases
Location SIP Servers –RFC 3261
Network Load Balancing
Unit OS10: Fault Tolerance
Load Weighting and Priority
Auburn University COMP7500 Advanced Operating Systems I/O-Aware Load Balancing Techniques (2) Dr. Xiao Qin Auburn University.
The Google File System Sanjay Ghemawat, Howard Gobioff and Shun-Tak Leung Google Presented by Jiamin Huang EECS 582 – W16.
Process Migration Troy Cogburn and Gilbert Podell-Blume
MapReduce: Simplified Data Processing on Large Clusters
Presentation transcript:

Automatic Load Balancing of ProActive Applications... with Fault Tolerance Javier Bustos, José Piquer, Universidad de Chile, Santiago, Chile Sebastien Boukhalfa, University Nice Sophia Antipolis Wednesday November th, 2006

Active Objects Migration migrateTo(…)

Dynamic Load Balancing 2 mechanisms working in parallel: Sender initiated: a quick reaction against overloading Receiver initiated: clustering active objects on the best machines

Sender Initiated

Receiver Initiated

Fault Tolerance Periodic record of active object states and incomming request (from the request queue) = checkpointing Using FT servers to store the information.

Union of ¿disjoint? schemes DYNAMIC LOAD BALANCING FAULT TOLERANCE FORWARDER

What is a forwarder? migrateTo(…) forwarder Active Object is not here anymore, it’s there 

… a forwarder Is not stored on Fault Tolerance servers because it is stateless (is not a part of the application, is part of the infrastructure) Has limited time to live Is created after each migration

Forwarder problem If the new location machine fails, how to update the forwarder? What if the last checkpoint is from the last location? If the forwarder machine fails, how to update the active object references? forwarder

… a forwarder Is not stored on Fault Tolerance servers because it is stateless (is not a part of the application, is part of the infrastructure) Has limited time to live Is created after each migration Update the references … but we have a FTServer which can do that!

Solution… Store the location of active objects on the FT Server, updating the location after each migration Using a hybrid scheme: forwarders and migration server (the FT server) “in case of an inconsistent reference, please consult the FT Server”

Conclusions and Future Work Dynamic load balancing and fault tolerance can live together Forwarder problem solved using the FT infrastructure of ProActive Future Work: Using the Dynamic Load Balancing infrastructure to find a suitable location after a fault.