Platform Support of Autonomic Computing: an Evolution of Manageability Architecture

Abstract

Autonomic Computing (AC) is maturing from a design philosophy to an emerging set of technologies and products that addresses the complexity of managing today's heterogeneous data centers and computing environments. This overview paper explains our motivation and outlines our technologies that provide platform support for AC. Specifically, we are developing platforms with sufficient support and on-board intelligence to enable autonomic capabilities, such as selfhealing and self-protecting, as well as features such as discovery and asset tracking, even when the host Operating System (OS) is inactive. To achieve these ends we are dedicating select platform resources and firmware, both exposed via well-defined standard interfaces, to implement a set of management and autonomic capabilities, and in the future, we hope to extend these platform autonomic capabilities, with appropriate management policies, to groups of platforms and eventually to the entire data center. Such interconnected autonomic platforms will provide the infrastructure and fabric to support service-oriented, grid, and utility computing. This paper also expounds on the Intel ® Active Management Technology † (Intel AMT) which is the first product incarnation of a framework and dedicated platform execution environment for AC. We also discuss how manageability architectures need to evolve to support autonomic behavior at a group level, such as defining interactions among platforms within a group to collectively deliver on specific goals and implement group-level policies. We cite examples related to malware detection and power management that illustrate how this new approach to managing IT infrastructure works. INFORMATION TECHNOLOGY OVERVIEW Today, Information Technology (IT) departments are plagued by increasing complexity, poor utilization of assets, space, and high power consumption. Further, system management in such environments is fragmented and labor-intensive requiring considerable human involvement for mundane functions such as configuration and provisioning. Such common inefficiencies in data centers force IT managers to provision their data centers for peak loads resulting in low average resource utilization and availability in the range of 99.99% or possibly less. Current data center deployments tend to have strong and static binding between servers and applications, as well as between servers and administrators, due mainly to a lack of technology to deploy otherwise. To improve the efficiency of assets, space, and power, IT managers turned to virtualization, which provides the ability to consolidate multiple applications onto a single server (a server could be stand-alone, rack-mounted or bladed). These servers may include advanced technologies that support multiple logical and physical partitions within each physical machine such as virtualization, partitioning hooks, and multiple cores. These advanced technologies have increased the cost and complexity of platforms and require that management solutions adapt to deal with them. Unfortunat...

Continuous Gossip-Based Aggregation through Dynamic Information Aging

“…While central monitoring approaches like the SNMP [3] or Intel's Active Management Technology (AMT) [4] support the monitoring of small-to mid-scale networks, e.g., the quality of service in the network and transport layer [5], there are several reasons why they are not suitable for selforganizing large-scale peer-to-peer networks that aim at full decentrality. Decentralized monitoring approaches (see Section II) are required, these are mainly either tree-based or gossipbased.…”