Joseph Phillip Bigus

Using control theory to achieve service level objectives in performance management

A widely used approach to achieving service level objectives for a software system (e.g., an email server) is to add a controller that manipulates the target systems tuning parameters. We describe a methodology for designing such controllers for software systems that builds on classical control theory. The classical approach proceeds in two steps: system identification and controller design. In system identification, we construct mathematical models of the target system. Traditionally, this has been based on a first-principles approach, using detailed knowledge of the target system. Such models can be complex and difficult to build, validate, use, and maintain. In our methodology, a statistical (ARMA) model is fit to historical measurements of the target being controlled. These models are easier to obtain and use and allow us to apply control-theoretic design techniques to a larger class of systems. When applied to a Lotus Notes groupware server, we obtain model-fits with R2 no lower than 75% and as high as 98%. In controller design, an analysis of the models leads to a controller that will achieve the service level objectives. We report on an analysis of a closed-loop system using an integral control law with Lotus Notes as the target. The objective is to maintain a reference queue length. Using root-locus analysis from control theory, we are able to predict the occurrence (or absence) of controller-induced oscillations in the systems response. Such oscillations are undesirable since they increase variability, thereby resulting in a failure to meet the service level objective. We implement this controller for a real Lotus Notes system, and observe a remarkable correspondence between the behavior of the real system and the predictions of the analysis. This indicates that the control theoretic analysis is sufficient to select controller parameters that meet the desired goals, and the need for simulations is reduced.

ABLE: a toolkit for building multiagent autonomic systems

This paper describes a toolkit for building multiagent autonomic systems. The IBM Agent Building and Learning Environment (ABLE) provides a lightweight Java™ agent framework, a comprehensive JavaBeans™ library of intelligent software components, a set of development and test tools, and an agent platform. We describe a series of agents built using ABLE components and present three case studies of applications using the ABLE toolkit. The Autotune agent is a closed-loop controller agent that supports hierarchical distributed control. The Subsumption agent defines specific behaviors or strategies and can be plugged into a multiagent subsumption infrastructure. The Autonomic agent architecture features sensors and effectors for interacting with the external environment, layers of reflexive, reactive, and adaptive subsumption agents, components that dynamically model the autonomic system itself and its environment, and components for emotions, planning, and executive-level decision-making. By using the ABLE component library to build agents running on the ABLE distributed agent platform, we discuss how we can incrementally add new behaviors and capabilities to intelligent, autonomic systems.

Managing Web server performance with AutoTune agents

Managing the performance of e-commerce sites is challenging. Site content changes frequently, as do customer interests and business plans, contributing to dynamically varying workloads. To maintain good performance, system administrators must tune their information technology environment on an ongoing basis. Unfortunately, doing so requires considerable expertise and increases the total cost of system ownership. In this paper, we propose an agent-based solution that not only automates the ongoing system tuning but also automatically designs an appropriate tuning mechanism for the target system. We illustrate this in the context of managing a Web server. There we study the problem of controlling CPU and memory utilization of an Apache® Web server using the application-level tuning parameters MaxClients and KeepAlive, which are exposed by the server. Using the AutoTune agent framework under the Agent Building and Learning Environment (ABLE), we construct agents to fully automate a control-theoretic methodology that involves model building, controller design, and run-time feedback control. Specifically, we design (1) a modeling agent that builds a dynamic system model from the controlled server run data, (2) a controller design agent that uses optimal control theory to derive a feedback control algorithm customized to that server, and (3) a run-time control agent that deploys the feedback control algorithm in an on-line real- time environment to automatically manage the Web server. The designed autonomic feedback control system is able to handle the dynamic and interrelated dependencies between the1 tuning parameters and the performance metrics with guaranteed stability from control theory. The effectiveness of the AutoTune agents is demonstrated through experiments involving variations in workload, server capacity, and business objectives. The results also serve as a validation of the ABLE toolkit and the AutoTune agent framework.

SWORD: scalable and flexible workload generator for distributed data processing systems

Workload generation is commonly employed for performance characterization, testing and benchmarking of computer systems and networks. Workload generation typically aims at simulating or emulating traffic generated by different types of applications, protocols and activities, such as Web browsing, email, chat, as well as stream multimedia traffic. We present a scalable workload generator (SWORD) that we have developed for the testing and benchmarking of high-volume data processing systems. The tool is not only scalable but is also flexible and extensible allowing the generation of workload of a variety of types of applications and of contents

Bringing planning to autonomic applications with ABLE

Planning has received tremendous interest as a research area within AI over the last three decades but it has not been applied commercially as widely as its other AI counterparts like learning or data mining. The reasons are many: the utility of planning in business applications was unclear, the planners used to work best in small domains and there was no general purpose planning and execution infrastructure widely available. Much has changed lately. Compelling applications have emerged, e.g., computing systems have become so complex that the IT industry recognizes the necessity of deliberative methods to make these systems self-configuring, self-healing, self-optimizing and self-protecting. Planning has seen an upsurge in the last decade with new planners that are orders of magnitude faster than before and are able to scale this performance to complex domains, e.g., those with metric and temporal constraints. However, planning and execution infrastructure is still tightly tied to a specific application which can have its own idiosyncrasies. In this paper, we fill the infrastructural gap by providing a domain independent planning and execution environment that is implemented in the ABLE agent building toolkit, and demonstrate its ability to solve practical business applications. The planning-enabled ABLE is publicly available and is being used to solve a variety of planning applications in IBM including the self-management/autonomic computing scenarios.

The agent building and learning environment

This paper describes the Agent Building and Learning Environment (ABLE) a Java-based framework for developing and deploying hybrid intelligent agents and agent applications. ABLE provides a set of reusable JavaBean components, called AbleBeans, along with several flexible interconnection methods for combining those components to create software agents. AbleBeans implement data access, filtering and transformation, learning, and reasoning capabilities. Function-specific AbleAgents are provided for classification, clustering, prediction, and genetic search. Application-specific agents can be constructed using one or more of these AbleBeans. AbleAgents are situated in their environment through the use of sensors and effectors, which provide a generic mechanism for linking them to Java applications. A GUI-based interactive development environment, the Able Agent Editor, is provided to assist in the construction of AbleAgents using AbleBean components. The Able agent platform is a FIPA-compliant distributed framework for creating multi-agent systems. The utility of the ABLE framework has been proven through its use in several IBM

Information technology for healthcare transformation

Rising costs, decreasing quality of care, diminishing productivity, and increasing complexity have all contributed to the present state of the healthcare industry. The interactions between payers (e.g., insurance companies and health plans) and providers (e.g., hospitals and laboratories) are growing and are becoming more complicated. The constant upsurge in and enhanced complexity of diagnostic and treatment information has made the clinical decision-making process more difficult. Medical transaction charges are greater than ever. Population-specific financial requirements are increasing the economic burden on the entire system. Medical insurance and identity theft frauds are on the rise. The current lack of comparative cost analytics hampers systematic efficiency. Redundant and unnecessary interventions add to medical expenditures that add no value. Contemporary payment models are antithetic to outcome-driven medicine. The rate of medical errors and mistakes is high. Slow inefficient processes and the lack of best practice support for care delivery do not create productive settings. Information technology has an important role to play in approaching these problems. This paper describes IBM Researchs approach to helping address these issues, i.e., the evidence-based healthcare platform.

A framework for evidence-based health care incentives simulation

We present a general simulation framework designed for modeling incentives in a health care delivery system. This first version of the framework focuses on representing provider incentives. Key framework components are described in detail, and we provide an overview of how data-driven analytic methods can be integrated with this framework to enable evidence-based simulation. The software implementation of a simple simulation model based on this framework is also presented.

Applying a framework for healthcare incentives simulation

At WinterSim 2011, we originally proposed an agent-based framework for healthcare simulations, enabling flexible integration of multiple simulation models, including models of disease progression, effects of provider interventions, and provider behavior models that are responsive to contractual incentives. In this paper, we report results using our proposed framework to integrate two examples of provider behavior models, two examples of disease models, and four examples of payment models. We explore multiple combinations of these models and simulate the impact that alternative payment models may have on health and financial outcomes. These examples test the robustness of the simulation framework, and illustrate the value of such simulations to the policy makers who design incentives to improve cost and health outcomes, and to providers who wish to evaluate the financial impact of proposed incentives on their practice.