Feedback Control of Computing Systems

Reference: Hellerstein, J. L., Diao, Y., Parekh, S. & Tilbury, D. M. (2004). Feedback Control of Computing Systems. Wiley–IEEE Press. ISBN 978-0-471-26637-2. URL

Summary

This book is the standard text that brings classical control theory to bear on computing systems, and it is the reference source for the control-theoretic strand of self-adaptation. Its thesis is that the performance and resource management of software systems — web servers, databases, middleware, queues — can and should be engineered with the same feedback-control apparatus used in mechanical and electrical engineering: model the system, design a controller, and obtain provable guarantees about its closed-loop behaviour rather than tuning by trial and error.

The authors develop the toolkit systematically: black-box system identification to fit dynamic models (e.g. difference-equation / transfer-function models) to measured input–output data from a computing system; controller design (proportional, integral, PID and pole-placement techniques); and analysis of the resulting closed loop through the properties that name the book’s recurring acronym SASO — Stability, Accuracy, Settling time, and Overshoot. Worked examples target real systems such as Apache HTTP server tuning, Lotus Notes/email server admission control, and load and queue-length regulation. For the self-adaptive-systems community the book supplies the rigorous underpinning that papers like Engineering Self-Adaptive Systems through Feedback Loops invoke: when the MAPE-K loop is treated as a genuine control loop, this is where the design-and-guarantee machinery comes from.

Key Ideas

Applies classical (mostly linear) control theory to computing systems for performance and resource management.
System identification: fit empirical dynamic models to measured input/output data of a running system (black-box modelling).
Controller design: proportional/integral/PID and pole-placement controllers for computing targets.
SASO closed-loop properties — Stability, Accuracy, Settling time, Overshoot — as the analysis vocabulary.
Concrete case studies (Apache, email/Notes servers, queueing/admission control) demonstrating measured, guaranteed regulation.

Connections

Conceptual Contribution

Claim: The dynamic behaviour of computing systems can be managed with classical feedback control, yielding provable closed-loop guarantees (stability, accuracy, settling time, overshoot) instead of heuristic tuning.
Mechanism: Identify an empirical dynamic model of the target system, design a proportional/integral/PID or pole-placement controller against it, and analyse the closed loop with the SASO properties; validate on real server/queue case studies.
Concepts introduced/used: Control Theory, Feedback Loop, System Identification, MAPE-K
Stance: foundational textbook / methodology
Relates to: The rigorous control-theoretic basis behind the “loop as first-class artefact” stance of Engineering Self-Adaptive Systems through Feedback Loops and the automated controller synthesis of Automated Design of Self-Adaptive Software with Control-Theoretical Formal Guarantees.

Backlinks

Linked Pages

Automated Design of Self-Adaptive Software with Control-Theoretical Formal Guarantees

Reference: Filieri, A., Hoffmann, H. & Maggio, M. (2014). Automated Design of Self-Adaptive Software with Control-Theoretical Formal Guarantees. In Proc. 36th International Conference on Software Engineering (ICSE ’14), ACM, pp. 299–310. DOI: 10.1145/2568225.2568272. URL

Summary

This paper attacks a practical obstacle to using control theory for self-adaptation: designing a controller normally requires control-engineering expertise that most software engineers lack. The authors propose an automated method that, given a software system exposing a tunable knob and a measurable goal, builds a dynamic model of the system by lightweight experimentation and then synthesises a controller for it — turning controller design into a push-button step of the software process rather than a bespoke modelling exercise.

The technique performs on-line system identification to fit a simple (linear, adaptive) model relating the control input to the measured output, then automatically derives a controller whose parameters are chosen to deliver formal control-theoretic guarantees: stability of the closed loop, absence of steady-state error (the goal is met in the limit), bounded settling time, and robustness to modelling inaccuracy and disturbances. Because the model is re-estimated on-line, the controller adapts as the system or environment drifts. The result is a general, reusable way to give self-adaptive software the kind of provable regulation guarantees advocated by Feedback Control of Computing Systems and Engineering Self-Adaptive Systems through Feedback Loops, without demanding that each developer become a control engineer.

Key Ideas

Automated controller synthesis: given a knob and a measurable goal, generate a controller with no manual control-theory design.
On-line system identification fits a simple adaptive model of input→output behaviour and re-estimates it as the system drifts.
Delivers formal guarantees — stability, zero steady-state error (goal convergence), bounded settling time, robustness to disturbance/model error.
Makes control-theoretic adaptation reusable and accessible to software engineers, not just control specialists.
A concrete bridge from classical control theory to the MAPE-K loop’s Plan/Execute stages.

Connections

Conceptual Contribution

Claim: Control-theoretic self-adaptation can be automated — a controller with formal stability and convergence guarantees can be synthesised from a system’s own measured behaviour, removing the control-engineering expertise barrier.
Mechanism: On-line system identification fits an adaptive linear model of the control input’s effect on the goal metric; a controller is then derived automatically to guarantee stability, zero steady-state error, bounded settling time, and robustness, and is re-tuned as the model is re-estimated.
Concepts introduced/used: Control Theory, System Identification, Feedback Loop, MAPE-K, Runtime Assurance
Stance: engineering / method
Relates to: Operationalises the control-theoretic programme of Feedback Control of Computing Systems and Engineering Self-Adaptive Systems through Feedback Loops, and supplies one form of the runtime assurance called for by Software Engineering for Self-Adaptive Systems - A Second Research Roadmap.

Summary

This chapter makes a single, influential argument: the feedback loop should be a first-class, explicit entity in the engineering of self-adaptive software. The authors observe that although virtually every self-adaptive system is, at heart, one or more control loops, those loops are usually hidden — dispersed through the code and left implicit — which makes the system hard to understand, analyse, and assure. Their prescription is to surface the loop as an architectural element with named, documented parts, drawing directly on the long tradition of control engineering, where feedback loops are the primary object of design and analysis.

The paper anatomises the loop into its canonical activities — collect (sense), analyse, decide, and act — closely mirroring MAPE-K, and stresses the properties engineers must reason about by analogy to control theory: stability, accuracy, settling time, and overshoot, together with the handling of disturbances and the cost of sensing and actuation. It further argues for making the loop’s control properties explicit design concerns and for recognising when a system needs multiple, possibly interacting or hierarchical loops. By naming the feedback loop as the unit of design and importing control-theoretic vocabulary, the chapter became a standard reference for the “the loop is the architecture” stance in self-adaptive systems.

Key Ideas

Make the feedback loop first-class: name it, document its parts, and treat it as the primary architectural element — not logic buried in code.
Canonical loop activities — collect → analyse → decide → act — align with the MAPE-K phases.
Import control-engineering properties as design concerns: stability, accuracy, settling time, overshoot, disturbance rejection.
Reason explicitly about multiple and interacting loops (hierarchical or coordinated control).
Positions control theory as a source of both design guidance and assurance for adaptation.

Connections

Conceptual Contribution

Claim: Because a self-adaptive system is a feedback control system, the feedback loop should be an explicit, first-class engineering artefact whose control properties are designed and analysed, not an implicit by-product of the code.
Mechanism: Surface the loop as an architectural element with named collect/analyse/decide/act parts (the MAPE-K phases), reason about it with control-theoretic properties (stability, accuracy, settling time, overshoot), and structure systems as one or more coordinated loops.
Concepts introduced/used: Feedback Loop, MAPE-K, Control Theory, Self-Adaptation
Stance: engineering / position
Relates to: Deepens the feedback-loop theme of the SEAMS roadmaps (Software Engineering for Self-Adaptive Systems - A Research Roadmap); its control-theoretic programme is made rigorous in Feedback Control of Computing Systems and Automated Design of Self-Adaptive Software with Control-Theoretical Formal Guarantees.

In this vault

System Identification

(page does not exist)

Feedback Loop

A control structure in which a system observes the effect of its own actions (or of environmental change) and feeds that observation back to adjust subsequent behaviour toward a goal. Feedback control is the engineering backbone of self-adaptation: the MAPE-K loop is a feedback loop whose sensor readings (Monitor) close back onto the actions taken through effectors (Execute).

Making the feedback loop a first-class, visible architectural element — rather than logic buried in application code — is a recurring theme of the self-adaptive-systems literature and connects the field to classical control theory (stability, settling time, overshoot) as a source of design and assurance techniques.

In this vault

Control Theory

The engineering discipline of designing feedback controllers that drive a system toward a reference goal despite disturbances, and of analysing the resulting closed loop for properties such as stability, accuracy, settling time, and overshoot (the “SASO” properties). In self-adaptive software it is the rigorous underpinning of the Feedback Loop / MAPE-K stance: if an adaptive system is really a control loop, control theory supplies both the design methods and the guarantees.

The line runs from Feedback Control of Computing Systems (Hellerstein et al. 2004), which imports classical control into computing-systems performance management via system identification and controller design, through Engineering Self-Adaptive Systems through Feedback Loops (Brun et al. 2009), which argues the loop should be a first-class artefact, to Automated Design of Self-Adaptive Software with Control-Theoretical Formal Guarantees (Filieri, Hoffmann & Maggio 2014), which automates controller synthesis so software engineers get formal guarantees without doing the control-theory design by hand.

In this vault

The Vision of Autonomic Computing

Reference: Kephart, J. O. & Chess, D. M. (2003). The Vision of Autonomic Computing. IEEE Computer 36(1): 41–50. DOI: 10.1109/MC.2003.1160055. URL

Summary

This manifesto argues that the chief obstacle to further progress in computing is not raw capability but complexity: the installation, configuration, tuning, and maintenance of large heterogeneous systems is outstripping the ability of human administrators to manage them. Taking the human autonomic nervous system as its metaphor — it regulates heartbeat and temperature without conscious attention — the paper proposes autonomic computing: systems that manage themselves against high-level objectives supplied by administrators, who say what is wanted rather than how to achieve it.

The paper’s central architectural abstraction is the autonomic element: an autonomic manager wrapping a managed element, coupled to it through sensors and effectors, and running a closed control loop. That loop — Monitor, Analyse, Plan, Execute over a shared Knowledge base — is the MAPE-K loop that became the field’s organising pattern. Autonomic elements are composed into larger systems whose global self-management emerges from local manager behaviour and inter-element relationships (service agreements, negotiation).

Kephart and Chess frame the field’s goals through four self- properties*: self-configuration (systems configure themselves from high-level policy), self-healing (detect, diagnose, and repair faults), self-optimisation (continually tune performance and resource use), and self-protection (defend against and anticipate attacks). They lay out a research agenda spanning the individual-element level (theory of autonomic elements, lifecycle, learning) and the system level (emergent behaviour, negotiation, trust, and the engineering and assurance challenges of large adaptive systems).

Key Ideas

Complexity is the limiting factor; the answer is systems that manage themselves given high-level goals.
Autonomic element = autonomic manager + managed element, joined by sensors and effectors.
MAPE-K control loop: Monitor → Analyse → Plan → Execute over shared Knowledge.
Self-CHOP properties: self-configuration, self-healing, self-optimisation, self-protection.
Policy-based, goal-directed management: administrators specify objectives, not procedures.
System-level self-management as emergent from many interacting autonomic elements (agreements, negotiation, trust).
A broad research agenda naming learning, optimisation, negotiation, and assurance as open problems.

Connections

Conceptual Contribution

Claim: As systems outgrow human administrability, they must become self-managing — regulating their own configuration, health, performance, and defence against high-level objectives rather than step-by-step instructions.
Mechanism: Wrap each managed element in an autonomic manager that runs a Monitor–Analyse–Plan–Execute loop over a shared Knowledge base (MAPE-K), reading sensors and driving effectors; compose autonomic elements, coordinated by service agreements and negotiation, into self-managing systems judged by the self-CHOP properties.
Concepts introduced/used: Autonomic Computing, MAPE-K, Self-Adaptation, Feedback Loop, Models@run.time
Stance: vision / research agenda
Relates to: The origin point for the self-adaptive cluster — the architecture-based realisation in Rainbow - Architecture-Based Self-Adaptation with Reusable Infrastructure, the reference architecture of Self-Managed Systems - an Architectural Challenge, and the taxonomy of Self-Adaptive Software - Landscape and Research Challenges all build on this framing. The biological metaphor complements the biology-inspired adaptation in Self-Adaptation Self-Expression Self-Awareness ASCENS.

Self-Adaptation

A system’s capability to modify its own behaviour or structure at runtime in response to internal or environmental change, without external intervention. It is typically implemented as a MAPE-K loop (Monitor-Analyse-Plan-Execute over a Knowledge base), the control abstraction introduced by Autonomic Computing.

Feedback Control of Computing Systems

Summary

Key Ideas

Connections

Conceptual Contribution

Tags

Backlinks