Related Work on Autonomic Management AUTONOMIA

1
Related Work on Autonomic Management- AUTONOMIA -
Date October 17th 2007 Joon-Myung Kang,
Chang-Keun Park eliot, pck1982_at_postech.ac.kr DP
NM Lab., Dept. of CSE, POSTECH
2
Autonomic Management System

• Autonomia (Joon-Myung)
• High Performance Distributed Computing Laboratory
  (HPDC), University of Arizona (Director Dr.
  Salim Hariri)
• Xiangdong Dong, Salim Hariri, Lizhi Xue, Huoping
  Chen, Ming Zhang, Sathija Pavuluri, Soujanya Rao,
  AUTONOMIA An Autonomic Computing Environment,
  Proceedings of the Performance, Computing, and
  Communications Conference (IPCCC 2003)
• AutoMate (Chang-Keun)
• The Applied Software Systems Laboratory (TASSL),
  Rutgers University (Director Dr. Manish
  Parashar, his Ph.D advisor was Dr. Hariri at
  Syracuse University)
• M. Agrawal, V. Bhat, H.Liu, V. Matossian, V.
  Putty, C. Schmidt, G. Zhang, M.Parashar, B.
  Khargharia, S. Hariri, Automate Enabling
  Autonomic Applications On the Grid, Proceedings
  of Active Middleware Services (AMS 2003)

3
Autonomic Management System

• FOCALE (Joon-Myung)
• Motorola Labs.
• John Strassner
• John Strassner, Nazim Agoulmine, Elyes Lehtihet,
  FOCALE A Novel Autonomic Networking
  Architecture, Proceedings of Latin American
  Autonomic Computing Symposium (LACCS2006)
• Autonomic Service Architecture (Chang-Keun)
• Network Architecture Lab., University of Toronto
  (Director Dr. Alberto Leon-Garcia)
• Ramy Farha, Myung Sup Kim, Alberto Leon-Garcia,
  James-Won-Ki Hong, Towards an Autonomic Service
  Architecture, Proceedings of IP Operation and
  Management (IPOM2005)

4
AUTONOMIA An Autonomic Computing
EnvironmentXiangdong Dong, Salim Hariri, Lizhi
Xue, Huoping Chen, Ming Zhang, Sathija Pavuluri,
Soujanya RaoProceedings of the Performance,
Computing, and Communications Conference (IPCCC
2003)
5
Autonomia

• An Autonomic Computing Environment
• University of Arizona
• High Performance Distributed Computing Laboratory
• http//www.ece.arizona.edu/hpdc
• Director Salim Hariri
• Research
• Autonomic Power and Performance Management for
  Large-Scale Data Centers
• AUTONOMIA An Autonomic Computing Environment
• Self-Configuring
• Self-Healing
• Autonomous Middleware for Wireless Sensors
• Autonomic Computing Middleware
• Anomaly Detection and Prevention System (ADAPSY)

6
Introduction

• Unmanageable and Insecure System Complexity
• Explosion growth in information and integration
  technology (billions of devices)
• Ubiquitous Access to information through PCs,
  PDAs, Cells, smart appliances, etc. (millions of
  users)
• Severe shortage in skilled IT workers (hundreds
  of thousands in US) will increase 100 in the
  next six years
• Bottom Line
• The increasing system complexity is reaching a
  level beyond human ability to manage and secure

7
Autonomic Computing

• Its time to design and build computing systems
  capable of running themselves, adjusting to
  varying circumstances, and preparing their
  resources to handle most efficiently the
  workloads we put upon them. They must anticipate
  needs and allow users to concentrate on what they
  want to accomplish
• Paul Horn, IBM

8
Autonomia

• The goal of this paper
• Develop an autonomic computing infrastructure
  that supports complete control and management
  service in the prototype system, which they refer
  to as Autonomia
• Autonomia An autonomic control and management
• Provide dynamically programmable control and
  management services to support the development
  and deployment of smart (intelligent)
  applications
• Provide automated performance and fault tolerant
  support for smart applications
• Provide automated deployment, registration,
  discovery of components
• Provide automated configuration of applications
  and system resources
• Provide secure, open computing environment
• Autonomia environment
• Provides application developers with all the
  tools required to specify the appropriate control
  and management schemes, deploy, and configure the
  required software and hardware resources

9
Application Management Editor
Users Application
AME Parser
AIK Repository
AMS
Event Server
Autonomic Run-time System
Self Deploying
Application Management Repository
Application State Repository
Self Protecting
Self Optimizing
Self Healing
Policy Engine
Component Repository
Resource Repository
Application Delegated Manager (ADM)
Repository Interface
Host B
Host A
CS2
CS1
MAS A
MAS A
Task Agent
Resource Monitor
Task Agent
Resource Monitor
Component
Interface
Component
Interface
AMS Autonomic Middleware Service
AIK Application Information Knowledgebase
MAS Mobile Agent System
Agent Entry
Component Service Item
CS Computer Server
Node Entry
Component
Mobile Agent
MAE Service Item
Component Entry
10
Application Management Editor
Users Application
AME Parser
AMS
AIK Repository
Event Server
Autonomic Run-time System
Self Deploying
Application Management Repository
Application State Repository
Self Protecting
Self Optimizing
Self Healing
Policy Engine
Component Repository
Resource Repository
Application Delegated Manager (ADM)
Repository Interface
Host B
Host A
CS2
CS1
MAS A
MAS A
Task Agent
Resource Monitor
Task Agent
Resource Monitor
Component
Interface
Component
Interface
MAS Mobile Agent System
Agent Entry
Component Service Item
CS Computer Server
Node Entry
Component
Mobile Agent
MAE Service Item
Component Entry
11
Implementation

• Mobile Agent System
• Uniform execution environment independent of the
  underlying hardware architecture and operations
  system
• Based on Java/Jini technologies
• Agent Transport Protocol
• Application Management Editor
• Develop an application by selecting components
  from a well-defined library or libraries of
  components that are registered in the Component
  Repository
• Enables the user to specify the management
  requirements for each component that are needed
  to control and manage the required autonomic
  properties for that component
• Application Service Template (AST) is created and
  stored in the component repository

12
Implementation

• Autonomic Middleware Service
• Core autonomic management services required to
  achieve an autonomic computing services required
  to achieve an autonomic computing environment
• The services offered by the AMS are implemented
  using mobile agents
• The main services offered by AMS are Component
  Repository, Resource Repository, JavaSpaces, JMS,
  and the Fault and Security Performance Handler
• Each AMS service can be implemented using one or
  more algorithms that can be selected dynamically
  at runtime based on the current application state
  as well as the system state
• Component Repository
• Jini Lookup Service named Components
• A collection of components that are currently
  available to the users to develop their
  applications
• Components are implemented as Jini Services
• Resource Repository
• Jini Lookup Service named Resources
• It keeps track of all host registrations that are
  currently registered in the environment

13
Implementation

• Application Delegated Manager
• Plays the role of a broker between components and
  resources
• Has the responsibility of supervising the
  application execution at runtime
• Maintain an up-to-date list of resources as well
  as available components that can participate in
  executing a given application
• Select the set of resources that can meet the
  autonomic requirements of an application
• Dispatch the appropriate number of mobile agents
  to deploy, configure, perform on-line monitoring,
  analysis, and adaptation when the application
  cannot meet its requirements

14
Fault Recovery Mechanism

• Component Fault Recovery

• Step
• Checkpointing regularly
• Detect a component failure
• Report a component fault entry
• Notify the fault handler
• Read the component type
• Get another available MAS
• Dispatch an agent to the host
• Read the checkpoint
• Resume execution

15
Fault Recovery Example
16
Conclusions

• Summary
• A novel architecture to implement an autonomic
  computing environment (Autonomia)
• This makes the control and management of
  large-scale parallel and distributed applications
  autonomic
• It showed how they can achieve the self-healing
  when component, agents, and or systems fail
• Ongoing Work
• Currently implementing other autonomic attributes
  to make their environment self-optimizing and
  self-protecting

17
Application Management Editor

• AME
• Provides users with XML software tools to
  describe strategies that can be used to achieve
  the required autonomic properties
• Main functions of the editor are controlling the
  application editor workplace and storing the
  application management requirements in the
  component repository

BACK
18
Application Middleware Service

• AMS
• Dynamically control the allocated resources to
  maintain the application requirements during
  application execution.
• provides a common set of autonomic services
  (self-configuring, self-healing, self-protecting,
  self-defining, etc.)
• It is based on three procedures Monitoring,
  Analysis and Verification, and Adaptation
• Component Repository(CR)
• Jini Lookup Service (Components), collection of
  components that are currently available to the
  users to develop their applications
• Resource Repository(RR)
• Jini Lookup Service (Resources), It keeps track
  of all host registrations that are currently
  registered in the environment
• Control and Data Message Center
• JavaSpaces and Java Message Service
• the Fault and Security, Performance Handler

BACK
19
Application Delegated Manager

• ADM
• software agent responsible to configure, deploy,
  run and maintain the autonomic properties of the
  application at runtime
• Plays the role of a broker between components and
  resources
• maintain an up-to-date list of resources as well
  as available components that can participate in
  executing a given application
• select the set of resources that can meet the
  autonomic requirements of an application
• dispatch the appropriate number of mobile agents
  to deploy, configure, perform on-line monitoring,
  analysis, and adaptation when the application
  cannot meet its requirements.

BACK