Technology and Research
Intel® Technology Journal Home
Volume 10, Issue 04
Autonomic Computing
Table of Contents
Technical Reviewers
About This Journal
Intel Published Articles
Read Past Journals
Subscribe
E-Mail this Journal to a Colleague
Home  ›  Technology and Research  ›  Intel® Technology Journal  ›  Autonomic Computing
ITJ Autonomic Computing
Intel® Technology Journal
Featuring Intel's recent
research and development
 
Autonomic Computing
Volume 10    Issue 04    Published November 9, 2006
ISSN 1535-864X    DOI: 10.1535/itj.1004.03

  Section 4 of 10  
Standards for autonomic computing
Standards for autonomic computing

Computing systems of today are highly diverse, and this heterogeneous nature necessitates the interaction of autonomic managers with resources from multiple vendors. In order to complete the entire self-managing loop, it is imperative that components within a single autonomic manager be able to interact with each other. In a larger system with a hierarchical architecture, multiple autonomic managers must be able to interact with each other. The role of standards is invaluable in facilitating both these interactions.

Another related development in the computing industry is convergence on the concept of building applications and services using the principles of service orientation, often referred to as Service-Oriented Architecture (SOA). The use of Web services as a technology to build SOA systems has gained significant following in the industry. The leading motivator for the use of this technology is the ability to interoperate amongst implementations from multiple vendors. Given the significant investment, implementations, tools, and broad industry support for SOA and autonomic computing, it is both pragmatic and desirable to use the same or interoperable Web-service-based standards for both domains.



Figure 2: Standards for autonomic computing
click image for larger view
 

Starting with a single autonomic manager, we now examine the various touch points where standards should be defined, point out existing standards in that domain, and highlight gaps where they exist. Figure 2 depicts a standards stack for enabling autonomic computing spanning hardware management, OS/Application management, services, and business process management. In the sections below a representative example of standards from each layer is briefly described. Some descriptions pertain to more than one layer in the stack and have been included only once.

Protocols and data formats

All message exchanges between autonomic elements need to be built upon a common set of protocols and data formats. This section briefly describes the common protocols and data formats.

  • UDP, TCP/IP: IP (v4 and v6) is the most popular network layer protocol. Layered on IP, are the connection-oriented TCP and connectionless UDP protocols.
  • Hypertext Transfer Protocol, HTTP [28]: The most popular messaging protocol over TCP/IP is the use of HTTP (standardized by the W3C), which is a request/response protocol and which is the foundation of the World Wide Web.
  • Extensible Markup Language, XML [31]: XML is a simple text format used for the exchange of information and its use has facilitated the sharing of data across disparate systems in a machine-readable manner.

Messaging eventing and addressing

This section briefly describes the basic messaging and addressing standards.

  • SOAP [32]: SOAP is a lightweight message exchange protocol based on XML defined by the W3C for exchanging structured information in a decentralized distributed environment.
  • WS-Addressing [33]: WS-Addressing is a W3C recommendation and defines a transport neutral mechanism to address Web services and messages.
  • Additional details on standards in Figure 2 pertaining to Messaging, Eventing and Addressing are contained in the section "External Interfaces for Autonomic Computing Elements."

Resource sensing, monitoring and effecting (external interfaces)

At the lowest level of abstraction, is the ability to monitor and control resources using the appropriate external interfaces. Listed below are various standards that fall into this domain

  • Simple Network Management Protocol (SNMP): This is a network management standard from the Internet Engineering Task Force (IETF) [3]. It consists of a management data-model and a network protocol for querying the status of managed resources, executing remote commands and an eventing mechanism for dealing with asynchronous communication. It is widely adopted and includes an extensible management data model called a Management Information Base (MIB).
  • WS-Management/WSDM: WS-Management is a Web services protocol standardized by the DMTF for managing devices, services, and systems. It supports retrieval of object properties, enumeration of objects, and a publish/subscribe communication paradigm (eventing). Web Services Distributed Management (WSDM) [7] is a management standard from the Organization for the Advancement of Structured Information Standards (OASIS) which consists of two components, Management Using Web Services (MUWS) and Management of Web Services (MOWS). Additional details are given later in this paper.
  • Intelligent Platform Management Interface (IPMI): IPMI [11] defines a message-based interface to intelligent platform hardware and a set of standardized records for describing and accessing platform managed devices, such as thermal sensors and fans. Keeping in mind the use of a common modeling schema, the IPMI developers have published guidelines for mapping IPMI to CIM [12].
  • Software Monitoring: Monitoring of software components remains a fragmented space, in large part due to the diverse programming languages and runtimes. It is expected that with the advent of a common management protocol and a data model (such as CIM) this fragmentation will be reduced. There is, however, good support for retrieving software instrumentation (although not the same across different platforms) using technologies such as Java Management Extensions (JMX) [9], System.Management namespace in .Net [10].


  Section 4 of 10  

In this article
Abstract
Introduction
Conceptual architecture for multi-level autonomic computing
Standards for autonomic computing
Description
External interfaces for autonomic computing elements
Summary
Acknowledgments
References
Authors' biographies
Download a PDF of this article.    Email This Page
Back to Top