arXiv subjects

Alan Dearle

Publications and source records attributed to Alan Dearle.

32 records · Page 2Linked to original sources

arXiv2010

A Flexible and Secure Deployment Framework for Distributed Applications

This paper describes an implemented system which is designed to support the deployment of applications offering distributed services, comprising a number of distributed components. This is achieved by creating high level placement and topology descriptions which drive tools that deploy applications consisting of components running on multiple hosts. The system addresses issues of heterogeneity by providing abstractions over host-specific attributes yielding a homogeneous run-time environment into which components may be deployed. The run-time environments provide secure binding mechanisms that permit deployed components to bind to stored data and services on the hosts on which they are running.

cs.DC↗

arXiv2010

A Framework for Constraint-Based Deployment and Autonomic Management of Distributed Applications

We propose a framework for deployment and subsequent autonomic management of component-based distributed applications. An initial deployment goal is specified using a declarative constraint language, expressing constraints over aspects such as component-host mappings and component interconnection topology. A constraint solver is used to find a configuration that satisfies the goal, and the configuration is deployed automatically. The deployed application is instrumented to allow subsequent autonomic management. If, during execution, the manager detects that the original goal is no longer being met, the satisfy/deploy process can be repeated automatically in order to generate a revised deployment that does meet the goal.

cs.DC↗

arXiv2010

A Peer-to-Peer Middleware Framework for Resilient Persistent Programming

The persistent programming systems of the 1980s offered a programming model that integrated computation and long-term storage. In these systems, reliable applications could be engineered without requiring the programmer to write translation code to manage the transfer of data to and from non-volatile storage. More importantly, it simplified the programmer's conceptual model of an application, and avoided the many coherency problems that result from multiple cached copies of the same information. Although technically innovative, persistent languages were not widely adopted, perhaps due in part to their closed-world model. Each persistent store was located on a single host, and there were no flexible mechanisms for communication or transfer of data between separate stores. Here we re-open the work on persistence and combine it with modern peer-to-peer techniques in order to provide support for orthogonal persistence in resilient and potentially long-running distributed applications. Our vision is of an infrastructure within which an application can be developed and distributed with minimal modification, whereupon the application becomes resilient to certain failure modes. If a node, or the connection to it, fails during execution of the application, the objects are re-instantiated from distributed replicas, without their reference holders being aware of the failure. Furthermore, we believe that this can be achieved within a spectrum of application programmer intervention, ranging from minimal to totally prescriptive, as desired. The same mechanisms encompass an orthogonally persistent programming model. We outline our approach to implementing this vision, and describe current progress.

cs.DC↗

arXiv2010

RAFDA: A Policy-Aware Middleware Supporting the Flexible Separation of Application Logic from Distribution

Middleware technologies often limit the way in which object classes may be used in distributed applications due to the fixed distribution policies that they impose. These policies permeate applications developed using existing middleware systems and force an unnatural encoding of application level semantics. For example, the application programmer has no direct control over inter-address-space parameter passing semantics. Semantics are fixed by the distribution topology of the application, which is dictated early in the design cycle. This creates applications that are brittle with respect to changes in distribution. This paper explores technology that provides control over the extent to which inter-address-space communication is exposed to programmers, in order to aid the creation, maintenance and evolution of distributed applications. The described system permits arbitrary objects in an application to be dynamically exposed for remote access, allowing applications to be written without concern for distribution. Programmers can conceal or expose the distributed nature of applications as required, permitting object placement and distribution boundaries to be decided late in the design cycle and even dynamically. Inter-address-space parameter passing semantics may also be decided independently of object implementation and at varying times in the design cycle, again possibly as late as run-time. Furthermore, transmission policy may be defined on a per-class, per-method or per-parameter basis, maximizing plasticity. This flexibility is of utility in the development of new distributed applications, and the creation of management and monitoring infrastructures for existing applications.

cs.DC↗

arXiv2010

Towards Adaptable and Adaptive Policy-Free Middleware

We believe that to fully support adaptive distributed applications, middleware must itself be adaptable, adaptive and policy-free. In this paper we present a new language-independent adaptable and adaptive policy framework suitable for integration in a wide variety of middleware systems. This framework facilitates the construction of adaptive distributed applications. The framework addresses adaptability through its ability to represent a wide range of specific middleware policies. Adaptiveness is supported by a rich contextual model, through which an application programmer may control precisely how policies should be selected for any particular interaction with the middleware. A contextual pattern mechanism facilitates the succinct expression of both coarse- and fine-grain policy contexts. Policies may be specified and altered dynamically, and may themselves take account of dynamic conditions. The framework contains no hard-wired policies; instead, all policies can be configured.

cs.DC↗

arXiv2010

Promoting Component Reuse by Separating Transmission Policy from Implementation

In this paper we present a methodology and set of tools which assist the construction of applications from components, by separating the issues of transmission policy from component definition and implementation. This promotes a greater degree of software reuse than is possible using traditional middleware environments. Whilst component technologies are usually presented as a mechanism for promoting reuse, reuse is often limited due to design choices that permeate component implementation. The programmer has no direct control over inter-address-space parameter passing semantics: it is fixed by the distributed application's structure, based on the remote accessibility of the components. Using traditional middleware tools and environments, the application designer may be forced to use an unnatural encoding of application level semantics since application parameter passing semantics are tightly coupled with the component deployment topology. This paper describes how inter-address-space parameter passing semantics may be decided independently of component implementation. Transmission policy may be dynamically defined on a per-class, per-method or per-parameter basis.

cs.DC↗

arXiv2010

RAFDA: Middleware Supporting the Separation of Application Logic from Distribution Policy

Middleware technologies often limit the way in which object classes may be used in distributed applications due to the fixed distribution policies imposed by the Middleware system. These policies permeate the applications developed using them and force an unnatural encoding of application level semantics. For example, the application programmer has no direct control over inter-address-space parameter passing semantics since it is fixed by the application's distribution topology which is dictated early in the design cycle by the Middleware. This creates applications that are brittle with respect to changes in the way in which the applications are distributed. This paper explores technology permitting arbitrary objects in an application to be dynamically exposed for remote access. Using this, the application can be written without concern for its distribution with object placement and distribution boundaries decided late in the design cycle and even dynamically. Inter-address-space parameter passing semantics may also be decided independently of object implementation and at varying times in the design cycle, again, possibly as late as run-time. Furthermore, transmission policy may be defined on a per-class, per-method or per-parameter basis maximizing plasticity. This flexibility is of utility in the development of new distributed applications and the creation of management and monitoring infrastructures for existing applications.

cs.DC↗

arXiv2010

Orthogonal Persistence Revisited

The social and economic importance of large bodies of programs and data that are potentially long-lived has attracted much attention in the commercial and research communities. Here we concentrate on a set of methodologies and technologies called persistent programming. In particular we review programming language support for the concept of orthogonal persistence, a technique for the uniform treatment of objects irrespective of their types or longevity. While research in persistent programming has become unfashionable, we show how the concept is beginning to appear as a major component of modern systems. We relate these attempts to the original principles of orthogonal persistence and give a few hints about how the concept may be utilised in the future.

cs.PL↗

arXiv2010

Generating a Family of Byzantine Tolerant Protocol Implementations Using a Meta-Model Architecture

We describe an approach to modelling a Byzantine tolerant distributed algorithm as a family of related finite state machines, generated from a single meta-model. Various artefacts are generated from each state machine, including diagrams and source-level protocol implementations. The approach allows a state machine formulation to be applied to problems for which it would not otherwise be suitable, increasing confidence in correctness.

cs.DC↗

arXiv2010

Applying Constraint Solving to the Management of Distributed Applications

We present our approach for deploying and managing distributed component-based applications. A Desired State Description (DSD), written in a high-level declarative language, specifies requirements for a distributed application. Our infrastructure accepts a DSD as input, and from it automatically configures and deploys the distributed application. Subsequent violations of the original requirements are detected and, where possible, automatically rectified by reconfiguration and redeployment of the necessary application components. A constraint solving tool is used to plan deployments that meet the application requirements.

cs.DC↗

arXiv2010

Hosting Byzantine Fault Tolerant Services on a Chord Ring

In this paper we demonstrate how stateful Byzantine Fault Tolerant services may be hosted on a Chord ring. The strategy presented is fourfold: firstly a replication scheme that dissociates the maintenance of replicated service state from ring recovery is developed. Secondly, clients of the ring based services are made replication aware. Thirdly, a consensus protocol is introduced that supports the serialization of updates. Finally Byzantine fault tolerant replication protocols are developed that ensure the integrity of service data hosted on the ring.

cs.DC↗

arXiv2010

H2O: An Autonomic, Resource-Aware Distributed Database System

This paper presents the design of an autonomic, resource-aware distributed database which enables data to be backed up and shared without complex manual administration. The database, H2O, is designed to make use of unused resources on workstation machines. Creating and maintaining highly-available, replicated database systems can be difficult for untrained users, and costly for IT departments. H2O reduces the need for manual administration by autonomically replicating data and load-balancing across machines in an enterprise. Provisioning hardware to run a database system can be unnecessarily costly as most organizations already possess large quantities of idle resources in workstation machines. H2O is designed to utilize this unused capacity by using resource availability information to place data and plan queries over workstation machines that are already being used for other tasks. This paper discusses the requirements for such a system and presents the design and implementation of H2O.

cs.DB↗

arXiv2010

Autonomic Management of Maintenance Scheduling in Chord

This paper experimentally evaluates the effects of applying autonomic management to the scheduling of maintenance operations in a deployed Chord network, for various membership churn and workload patterns. Two versions of an autonomic management policy were compared with a static configuration. The autonomic policies varied with respect to the aggressiveness with which they responded to peer access error rates and to wasted maintenance operations. In most experiments, significant improvements due to autonomic management were observed in the performance of routing operations and the quantity of data transmitted between network members. Of the autonomic policies, the more aggressive version gave slightly better results.

cs.NI↗

arXiv2010

An Approach to Ad hoc Cloud Computing

We consider how underused computing resources within an enterprise may be harnessed to improve utilization and create an elastic computing infrastructure. Most current cloud provision involves a data center model, in which clusters of machines are dedicated to running cloud infrastructure software. We propose an additional model, the ad hoc cloud, in which infrastructure software is distributed over resources harvested from machines already in existence within an enterprise. In contrast to the data center cloud model, resource levels are not established a priori, nor are resources dedicated exclusively to the cloud while in use. A participating machine is not dedicated to the cloud, but has some other primary purpose such as running interactive processes for a particular user. We outline the major implementation challenges and one approach to tackling them.

cs.DC↗