The 2nd IEEE International Conference on
Autonomic Computing (ICAC-05)

Home

Organization

Submission

 Sponsors

Venue

Program

Tutorials

Previous Events

 

 Seattle, Washington
13 – 16 June, 2005

GENERAL CHAIRS

Manish Parashar, Rutgers Univ., USA 

Jeffrey Kephart, IBM Research, USA


PROGRAM CHAIRS

Karsten Schwan, Georgia Tech, USA 

Yi-Min Wang, Microsoft Research, USA


REGISTRATION

(ONLINE) (DOC) (PDF)


CALL FOR PAPERS

  Download in PDF format
  Download in DOC format
  Download in TXT format


CALL FOR TUTORIALS

  Download in PDF format
  Download in DOC format

 

CALL FOR DEMOS & EXHIBITIONS

  Download in PDF format
  Download in DOC format


IMPORTANT DATES

Title/Abstracts due:17:00 EST, Jan 17, 2005
Full papers due:17:00 EST, Jan 24, 2005

Author notification: Feb 28, 2005
Poster submissions: Mar 03, 2005
Tutorial submissions: Mar 03, 2005
Workshop submissions: Mar 03, 2005
Demo/Exhibit submission: Mar 03, 2005
Final manuscripts due: Apr 01, 2005
Conference: Jun 13-16, 2005

FURTHER INFORMATION

WWW: www.autonomic-conference.org
E-mail:
icac@caip.rutgers.edu

 

ICAC 2005 Industry Session

Chair: Karsten Schwan

(Tuesday, June 14, 04:00 - 05:30 PM )

Platform Support for Autonomic Computing
Milan Milenkovic, Lenitra Durham, Intel Labs 

Abstract:  Autonomic systems are generally regarded as addressing many issues in dealing with increasing complexity of managing computer systems and services.  In this presentation we discuss the value and benefits of platform support for autonomics - such as enriched instrumentation, standardized way of accessing sensors and effectors, and integration with management and policy tools.  As an illustration, we present an autonomic research platform designed to explore the nature and value of platform support for autonomic computing. The platform combines (1) hardware sensors, such as processor temperature and fan speed, with (2) software sensors – such as kernel memory usage and process queue length – and with (3) ambient sensors, based on wireless sensors (motes) placed in its surroundings. The platform contains a graphical authoring tool for creation of autonomic policies and behaviors that use and process sensory data to create actions on the managed platform expressed by manipulation of effectors.


Model-Driven Automation
Kumar Goswami, HP

Abstract:  In this talk, I will present some of the research we are doing at HP Labs and the holistic approach we are taking towards automating the lifecycle of applications. I will also discuss a specific engagement we have with HP-IT to highlight two things: 1) how we are using some of these technologies to automate a service utility and 2) the cultural road-blocks to the adoption of such technologies.


Autonomic Management using WSDM, an Interoperability Demonstration by IBM and HP
Heather Kreger, IBM, Bryan Murray, HP

Abstract: This demo will actually show give attendees a chance to get close to the Industry Session demonstrations by HP and IBM. The two demonstrations, created jointly by HP and IBM, to illustrate the value and interoperability of WSDM 1.0. WSDM was recently approved as a standard by the OASIS organization. Web Services Distributed Management is composed of two specifications, called MUWS (Management Using Web Services) and MOWS (Management Of Web Services). The MUWS specification, which is the foundation for bringing the power of Web services and SOA to management applications, is the center of the demonstrations. The two simple demos clearly illustrate how WSDM allows management applications to go beyond centralized monitoring in an agent-based world to distributed, collaborating management application in a world that is composed of both smart and dumb resources.

You may see the Weather Station Service Management demo, which illustrates how a client application is kept unaffected by the changes in status of the services it invokes. In this example, a weather client that depends on weather station services to send reports. WSDM allows a standard manager to dynamically discover the weather stations available, to get notified when one of the stations is taken down for maintenance and to automatically redirect the client application to the next best weather station available based on a cost policy. One of the interesting points illustrated in the demo, is the ability for a manager, the weather client manager, without administration rights on the resources, the Weather station services, to still access a subset of the manageability capabilities of the resource, such as registering for notifications and obtaining status, in order to manage the availability of its own resources, the clients. The resource could make capabilities available to each manager based on policy.

 Or you may see the Blackberry demo, which shows a resource constrained WSDM-enabled handheld device (in this case a Blackberry PDA) which is managed by a specialized autonomic Blackberry manager, which detects the installation of and removes unauthorized applications on the device for security reasons. This manager also acts as a virtual resource, presenting to the outside a view of a smarter and self-managed Blackberry that a higher-level manager is able support generically. This demo also illustrates the important point that the WSDM protocol is lightweight and can easily be implemented on resource-constrained devices. It also illustrates how different managers (produced by different companies, in this case IBM and HP) participate in an open and distributed management system based on WSDM, communicating with the resource as well as each other.

Both scenarios are made possible in a heterogeneous environment by the use of a common, standard, Web services-based management protocol, WSDM 1.0. The use of WSDM has many other benefits not illustrated in these particular examples that the demo presenters will be happy to discuss, such as the ability to enrich business processes with management information/action as well as the ability to instrument and manage across organizational boundaries.

©ICAC, 2005 All rights reserved