Second International Conference on Autonomic Computing (ICAC'05)

We consider pervasive computing applications that process and aggregate data-streams emanating from highly distributed data sources to produce a stream of updates that have an implicit business-value. Middleware that enables such aggregation of data-streams must support scalable and efficient self-management to deal with changes in the operating conditions and should have an embedded business-sens... View full abstract»

Grid environments enable users to share nondedicated resources that lack performance guarantees. This paper describes the design of application-centric middleware components to automatically recover from failures and dynamically adapt to grid environments with changing resource availabilities, improving fault-tolerance and performance. The key components of the application-centric approach are a g... View full abstract»

In this paper we present the design and evaluate the performance of an autonomic workflow execution engine. Although there exist many distributed workflow engines, in practice, it remains a difficult problem to deploy such systems in an optimal configuration. Furthermore, when facing an unpredictable workload with high variability, manual reconfiguration is not an option. Thanks to its autonomic c... View full abstract»

This paper presents an autonomic system in which two managers with different responsibilities collaborate to achieve an overall objective. The first, a node group manager, uses modeling and optimization algorithms to allocate server processes and individual requests among a set of server machines grouped into node groups, and also estimates its ability to fulfill its service-level objectives as a ... View full abstract»

When designing decentralised autonomic computing systems, a fundamental engineering issue is to assess system-wide behaviour. Such decentralised systems are characterised by the lack of global control, typically consist of autonomous cooperating entities, and often rely on self-organised emergent behaviour to achieve the requirements. A well-founded and practically feasible approach to study overa... View full abstract»

Scavenging (or resource borrowing) is a common approach used to harness unused resources to perform useful calculations. Since these are volunteer contributions from resource owners, it is vital to reduce the impact of scavenging activities on their native workload to a minimum. To this end, existing impact control systems are either overly conservative in stopping scavenging altogether or inflexi... View full abstract»

Increasingly, software systems are constructed by composing multiple existing applications. The resulting complexity increases the need for self-management of the system. However, adding autonomic behavior to composite systems is difficult, especially when the existing components were not originally designed to support such interactions. Moreover, entangling the code for integrated self-management... View full abstract»

Web applications suffer from software and configuration faults that lower their availability. Recovering from failure is dominated by the time interval between when these faults appear and when they are detected by site operators. We introduce a set of tools that augment the ability of operators to perceive the presence of failure: an automatic anomaly detector scours HTTP access logs to find chan... View full abstract»

We present an architecture for and prototype of a system for quickly detecting software problem recurrences. Re-discovery of the same problem is very common in many large software products and is a major cost component of product support. At run-time, when a problem occurs, the system collects the problem symptoms, including the program call-stack, and compares it against a database of symptoms to... View full abstract»

Detection and diagnosis of faults in a large-scale distributed system is a formidable task. Interest in monitoring and using traces of user requests for fault detection has been on the rise recently. In this paper we propose novel fault detection methods based on abnormal trace detection. One essential problem is how to represent the large amount of training trace data compactly as an oracle. Our ... View full abstract»

In this paper we present a new server monitoring method based on a new and powerful approach to dynamic data analysis: process query systems (PQS). PQS enables user-space monitoring of servers and, by using advanced behavioral models, makes accurate and fast decisions regarding server and service state. Data to support state estimation come from multiple sensor feeds located within a server networ... View full abstract»

The foremost crucial step towards a fully automated information lifecycle management (ILM) is to differentiate information by values in an unbiased manner and understand how values change over time. This paper presents an information valuation approach that quantifies the value of a given piece of information based on its usage over time. Our case study based on several real world NFS file server ... View full abstract»

Telecommunication systems have long enjoyed high levels of availability owing to the existence of a control plane. Software only solutions for server management have been available for many years, where observations are made on the server and processed on a central management station. Delegation of management decision making has not been fully realized. This paper describes the design of a true au... View full abstract»

Modern computing environments, such as enterprise data centers, Grids, and PlanetLab, introduce distributed services to address scalability, locality, and reliability. Web services (WS), in particular, improve decoupling, decentralization, and autonomicity within distributed systems. Unfortunately, scale and decentralization introduce additional problems in distributed services management, such as... View full abstract»

Methods for automatically managing the performance of computing services must estimate a performance model of that service. This paper explores properties that are necessary for performance model estimation of black-box computer systems when used together with adaptive feedback loops. It shows that the standard method of least-squares estimation often gives rise to models that make the control loo... View full abstract»

Policies are being increasingly used as a means of implementing autonomic computing features in IT systems. Since these policies consume system resources, the performance of a policy-based system may be hampered if resource intensive policies are scheduled at the same time when other policies and applications are being executed. However, with applications being accessed by users globally across ti... View full abstract»

If current trends in cellular phone technology, personal digital assistants, and wireless networking are indicative of the future, we can expect our environments to contain an abundance of networked computational devices and resources. We envision these devices acting in an orchestrated manner to meet users' needs, pushing the level of interaction away from particular devices and towards interacti... View full abstract»

Threshold violations reported for system components signal undesirable conditions in the system. In complex computer systems, characterized by dynamically changing workload patterns and evolving business goals, the pre-computed performance thresholds on the operational values of performance metrics of individual system components are not available. This paper focuses on a fundamental enabling tech... View full abstract»

Dynamic capacity provisioning is a useful technique for handling the multi-time-scale variations seen in Internet workloads. In this paper, we propose a novel dynamic provisioning technique for multitier Internet applications that employs (i) a flexible queuing model to determine how much resources to allocate to each tier of the application, and (ii) a combination of predictive and reactive metho... View full abstract»

Large data centers host several application environments (AEs) that are subject to workloads whose intensity varies widely and unpredictably. Therefore, the servers of the data center may need to be dynamically redeployed among the various AEs in order to optimize some global utility function. Previous approaches to solving this problem suffer from scalability limitations and cannot easily address... View full abstract»

High-end servers that can be partitioned into logical subsystems and repartitioned on the fly are now becoming available. This development raises the possibility of reconfiguring distributed systems online to optimize for dynamically changing workloads. This paper presents the initial steps towards a system that can learn to alter its current configuration in reaction to the current workload. In p... View full abstract»