Andres Quiroz

Towards autonomic workload provisioning for enterprise Grids and clouds

This paper explores autonomic approaches for optimizing provisioning for heterogeneous workloads on enterprise Grids and clouds. Specifically, this paper presents a decentralized, robust online clustering approach that addresses the distributed nature of these environments, and can be used to detect patterns and trends, and use this information to optimize provisioning of virtual (VM) resources. It then presents a model-based approach for estimating application service time using long-term application performance monitoring, to provide feedback about the appropriateness of requested resources as well as the systems ability to meet QoS constraints and SLAs. Specifically for high-performance computing workloads, the use of a quadratic response surface model (QRSM) is justified with respect to traditional models, demonstrating the need for application-specific modeling. The proposed approaches are evaluated using a real computing center workload trace and the results demonstrate both their effectiveness and cost-efficiency.

Energy-efficient application-aware online provisioning for virtualized clouds and data centers

As energy efficiency and associated costs become key concerns, consolidated and virtualized data centers and clouds are attractive computing platforms for data- and compute- intensive applications. These platforms provide an abstraction of nearly-unlimited computing resources through the elastic use of pools of consolidated resources, and provide opportunities for higher utilization and energy savings. Recently, these platforms are also being considered for more traditional high-performance computing (HPC) applications that have typically targeted Grids and similar conventional HPC platforms. However, maximizing energy efficiency, cost-effectiveness, and utilization for these applications while ensuring performance and other Quality of Service (QoS) guarantees, requires leveraging important and extremely challenging tradeoffs. These include, for example, the tradeoff between the need to efficiently create and provision Virtual Machines (VMs) on data center resources and the need to accommodate the heterogeneous resource demands and runtimes of these applications. In this paper we present an energy-aware online provisioning approach for HPC applications on consolidated and virtualized computing platforms. Energy efficiency is achieved using a workload-aware, just-right dynamic provisioning mechanism and the ability to power down subsystems of a host system that are not required by the VMs mapped to it. We evaluate the presented approach using real HPC workload traces from widely distributed production systems. The results presented demonstrated that compared to typical reactive or predefined provisioning, our approach achieves significant improvements in energy efficiency with an acceptable QoS penalty.

Peer-to-Peer Cloud Provisioning: Service Discovery and Load-Balancing

Clouds have evolved as the next-generation platform that facilitates creation of wide-area on-demand renting of computing or storage services for hosting application services that experience highly variable workloads and requires high availability and performance. Interconnecting Cloud computing system components (servers, virtual machines (VMs), application services) through peer-to-peer routing and information dissemination structure are essential to avoid the problems of provisioning efficiency bottleneck and single point of failure that are predominantly associated with traditional centralized or hierarchical approaches. These limitations can be overcome by connecting Cloud system components using a structured peer-to-peer network model (such as distributed hash tables (DHTs)). DHTs offer deterministic information/query routing and discovery with close to logarithmic bounds as regards network message complexity. By maintaining a small routing state of O (log n) per VM, a DHT structure can guarantee deterministic look-ups in a completely decentralized and distributed manner. This chapter presents: (i) a layered peer-to-peer Cloud provisioning architecture; (ii) a summary of the current state-of-the-art in Cloud provisioning with particular emphasis on service discovery and load-balancing; (iii) a classification of the existing peer-to-peer network management model with focus on extending the DHTs for indexing and managing complex provisioning information; and (iv) the design and implementation of novel, extensible software fabric (Cloud peer) that combines public/private clouds, overlay networking, and structured peer-to-peer indexing techniques for supporting scalable and self-managing service discovery and load-balancing in Cloud computing environments. Finally, an experimental evaluation is presented that demonstrates the feasibility of building next-generation Cloud provisioning systems based on peer-to-peer network management and information dissemination models. The experimental test-bed has been deployed on a public cloud computing platform, Amazon EC2, which demonstrates the effectiveness of the proposed peer-to-peer Cloud provisioning software fabric.

Design and Implementation of a Distributed Content-based Notification Broker for WS-Notification

We describe an implementation based on the WS-Notification (WSN) specification for publish/subscribe communication which provides a distributed, content-based notification service. The implementation is based on a distributed hashtable (DHT) built on a structured overlay of peer nodes. The entire system acts as a notification broker, so that notification producers and consumers that make use of the network can achieve loosely-coupled communication with a decentralized, scalable service. We develop and evaluate self-optimizing behavior built to reduce notification traffic within the network

Robust clustering analysis for the management of self-monitoring distributed systems

We present a decentralized algorithm for online clustering analysis used for anomaly detection in self-monitoring distributed systems. In particular, we demonstrate the monitoring of a network of printing devices that can perform the analysis without the use of external computing resources (i.e. in-network analysis). We also show how to ensure the robustness of the algorithm, in terms of anomaly detection accuracy, in the face of failures of the network infrastructure on which the algorithm runs. Further, we evaluate the tradeoff in terms of overhead necessary for ensuring this robustness and present a method to reduce this overhead while maintaining the detection accuracy of the algorithm.

Design and evaluation of decentralized online clustering

Ensuring the efficient and robust operation of distributed computational infrastructures is critical, given that their scale and overall complexity is growing at an alarming rate and that their management is rapidly exceeding human capability. Clustering analysis can be used to find patterns and trends in system operational data, as well as highlight deviations from these patterns. Such analysis can be essential for verifying the correctness and efficiency of the operation of the system, as well as for discovering specific situations of interest, such as anomalies or faults, that require appropriate management actions. This work analyzes the automated application of clustering for online system management, from the point of view of the suitability of different clustering approaches for the online analysis of system data in a distributed environment, with minimal prior knowledge and within a timeframe that allows the timely interpretation of and response to clustering results. For this purpose, we evaluate DOC (Decentralized Online Clustering), a clustering algorithm designed to support data analysis for autonomic management, and compare it to existing and widely used clustering algorithms. The comparative evaluations will show that DOC achieves a good balance in the trade-offs inherent in the challenges for this type of online management.

Towards energy-aware autonomic provisioning for virtualized environments

As energy efficiency and associated costs become key concerns, consolidated and virtualized data centers and clouds are attractive computing platforms for data- and compute-intensive applications. Recently, these platforms are also being considered for more traditional high-performance computing (HPC) applications. However, maximizing energy efficiency, cost-effectiveness, and utilization for these applications while ensuring performance and other Quality of Service (QoS) guarantees, requires leveraging important and extremely challenging tradeoffs. These include, for example, the tradeoff between the need to efficiently create and provision Virtual Machines (VMs) on data center resources and the need to accommodate the heterogeneous resource demands and runtimes of the applications that run on them. In this paper we propose an energy-aware online provisioning approach for HPC applications on consolidated and virtualized computing platforms. Energy efficiency is achieved using a workload-aware, just-right dynamic provisioning mechanism and the ability to power down subsystems of a host system that are not required by the VMs mapped to it. Our preliminary evaluations show that our approach can improve energy efficiency with an acceptable QoS penalty.

Clustering Analysis for the Management of Self-Monitoring Device Networks

The increasing computing and communication capabilities of multi-function devices (MFDs) have enabled networks of such devices to provide value-added services. This has placed stringent QoS requirements on the operations of these device networks. This paper investigates how the computational capabilities of the devices in the network can be harnessed to achieve self-monitoring and QoS management. Specifically, the paper investigates the application of clustering analysis for detecting anomalies and trends in events generated during device operation, and presents a novel decentralized cluster and anomaly detection algorithm. The paper also describes how the algorithm can be implemented within a device overlay network, and demonstrates its performance and utility using simulated as well as real workloads.

Autonomic policy adaptation using decentralized online clustering

Policies are a powerful means of expressing high-level, goal oriented parameters to manage the behavior of systems and users, and are thus valuable tools for autonomic management. However, the autonomic management of policies is in itself a challenging problem. Particularly, their applicability is typically limited in situations where management actions depend on dynamic system properties, which require adapting policy application thresholds and parameters without modifying absolute policy definition constraints. In this paper, we propose a novel policy definition framework that enables autonomic policy adaptation and proactive policy application, based on the online analysis and characterization of system operation and feedback events. Our approach is supported by a decentralized clustering mechanism and scalable distributed communication platform that together enable the online analysis of events and the efficient generation and enforcement of dynamic policies among distributed system components. We justify, with the evaluation of illustrative scenarios, the need for online data analysis for policy adaptation and the potential benefits of our approach.

A framework for distributed content-based web services notification in Grid systems

We describe a content-based distributed notification service for Publish/Subscribe communication that implements the WS-Notification (WSN) standards. The notification service is built on a messaging framework that supports an associative rendezvous messaging mechanism on a structured overlay of peer nodes. The entire system acts as a notification broker, so that notification publishers and subscribers outside the network can achieve loosely coupled communication through a decentralized, scalable service, by interacting with any of the broker peers. Self-optimizing mechanisms are built into the framework to reduce notification traffic within as well as from the peer network. We describe the framework and its performance and evaluate the effectiveness of the self-optimizing mechanisms.