Manish Marwah

Renewable and cooling aware workload management for sustainable data centers

Recently, the demand for data center computing has surged, increasing the total energy footprint of data centers worldwide. Data centers typically comprise three subsystems: IT equipment provides services to customers; power infrastructure supports the IT and cooling equipment; and the cooling infrastructure removes heat generated by these subsystems. This work presents a novel approach to model the energy flows in a data center and optimize its operation. Traditionally, supply-side constraints such as energy or cooling availability were treated independently from IT workload management. This work reduces electricity cost and environmental impact using a holistic approach that integrates renewable supply, dynamic pricing, and cooling supply including chiller and outside air cooling, with IT workload planning to improve the overall sustainability of data center operations. Specifically, we first predict renewable energy as well as IT demand. Then we use these predictions to generate an IT workload management plan that schedules IT workload and allocates IT resources within a data center according to time varying power supply and cooling efficiency. We have implemented and evaluated our approach using traces from real data centers and production systems. The results demonstrate that our approach can reduce both the recurring power costs and the use of non-renewable energy by as much as 60% compared to existing techniques, while still meeting the Service Level Agreements.

Minimizing data center SLA violations and power consumption via hybrid resource provisioning

This paper presents a novel approach to correctly allocate resources in data centers, such that SLA violations and energy consumption are minimized. Our approach first analyzes historical workload traces to identify long-term patterns that establish a “base” workload. It then employs two techniques to dynamically allocate capacity: predictive provisioning handles the estimated base workload at coarse time scales (e.g., hours or days) and reactive provisioning handles any excess workload at finer time scales (e.g., minutes). The combination of predictive and reactive provisioning achieves a significant improvement in meeting SLAs, conserving energy, and reducing provisioning costs. We implement and evaluate our approach using traces from four production systems. The results show that our approach can provide up to 35% savings in power consumption and reduce SLA violations by as much as 21% compared to existing techniques, while avoiding frequent power cycling of servers.

Probabilistic performance modeling of virtualized resource allocation

Virtualization technologies enable organizations to dynamically flex their IT resources based on workload fluctuations and changing business needs. However, only through a formal understanding of the relationship between application performance and virtualized resource allocation can over-provisioning or over-loading of physical IT resources be avoided. In this paper, we examine the probabilistic relationships between virtualized CPU allocation, CPU contention, and application response time, to enable autonomic controllers to satisfy service level objectives (SLOs) while more effectively utilizing IT resources. We show that with only minimal knowledge of application and system behaviors, our methodology can model the probability distribution of response time with a mean absolute error of less than 6% when compared with the measured response time distribution. We then demonstrate the usefulness of a probabilistic approach with case studies. We apply basic laws of probability to our model to investigate whether and how CPU allocation and contention affect application response time, correcting for their effects on CPU utilization. We find mean absolute differences of 8-10% between the modeled response time distributions of certain allocation states, and a similar difference when we add CPU contention. This methodology is general, and should also be applicable to non-CPU virtualized resources and other performance modeling problems.

TPC server fault tolerance using connection migration to a backup server

This paper describes the design, implementation, and performance evaluation of ST-TCP (Server fault-Tolerant TCP), which is an extension of TCP to tolerate TCP server failures. This is done by using an active backup server that keeps track of the state of the TCP connection and takes over the TCP connection whenever the primary fails. This migration of the TCP connection to the backup is completely transparent to the client. Because no changes are required on the client machine, any TCP client can access a ST-TCP server. The performance overhead of ST-TCP over standard TCP is minimal, and during normal operation its behavior is the same as that of a regular TCP. In addition, ST-TCP provides a fast and seamless failover whenever the primary server fails. This is verified by a prototype implementation of ST-TCP in the Linux operating system, and experiments with a number of simulated applications which have different communication characteristics.

Towards the design and operation of net-zero energy data centers

Reduction of resource consumption in data centers is becoming a growing concern for data center designers, operators and users. Accordingly, interest in the use of renewable energy to provide some portion of a data centers overall energy usage is also growing. One key concern is that the amount of renewable energy necessary to satisfy a typical data centers power consumption can lead to prohibitively high capital costs for the power generation and delivery infrastructure, particularly if on-site renewables are used. In this paper, we introduce a method to operate a data center with renewable energy that minimizes dependence on grid power while minimizing capital cost. We achieve this by integrating data center demand with the availability of resource supplies during operation. We discuss results from the deployment of our method in a production data center.

Data analysis, visualization and knowledge discovery in sustainable data centers

A significant amount of energy consumption is now attributed to data centers due to their ever increasing numbers, size and power densities. Thus, there are efforts focused at making a data center more sustainable by reducing its energy consumption and carbon footprint. This requires an end-to-end management approach with requirements derived from service level agreements (SLAs) and a flexible infrastructure that can be closely monitored and finely controlled. The infrastructure can then be manipulated to satisfy the requirements while optimizing for sustainability metrics and total cost of operations. In this paper, we explore the role of data analysis, visualization and knowledge discovery techniques in improving the sustainability of a data center. We present use cases from a large, sensor-rich, state-of-the-art data center on the application of these techniques to the three main sub-systems of a data center, namely, power, cooling and compute. Furthermore, we provide recommendations for where these techniques can be used within these sub-systems for improving sustainability metrics of a data center.

CHAOS: A Data Stream Analysis Architecture for Enterprise Applications

In this paper, we describe the design of our architecture for Continuous, Heterogeneous Analysis Over Streams, aka CHAOS that combines stream processing, approximation techniques, mining, complex event processing and visualization. CHAOS, with the novel concept of Computational Stream Analysis Cube, provides an effective, scalable platform for near real time processing of business and enterprise streams. We describe our approach with a real data center temperature analysis application.

Quantifying the sustainability impact of data center availability

Data center availability is critical considering the explosive growth in Internet services and peoples dependence on them. Furthermore, in recent years, sustainability has become important. However, data center designers have little information on the sustainability impact of data center availability architectures. In this paper, we present an approach to estimate the sustainability impact of such architectures. Availability is computed using Stochastic Petri Net (SPN) models while an exergy-based lifecycle assessment (LCA) approach is used for quantifying sustainability impact. The approach is demonstrated on real life data center power infrastructure architectures. Five different architectures are considered and initial results show that quantification of sustainability impact provides important information to a data center designer in evaluating availability architecture choices.

Towards an understanding of campus-scale power consumption

Commercial buildings are significant consumers of electricity. In this paper, we collect and analyze six weeks of data from 39 power meters in three buildings of a campus of a large company. We use an unsupervised anomaly detection technique based on a low-dimensional embedding to identify power saving opportunities. Further, to better manage resources such as lighting and HVAC, we develop occupancy models based on readily available port-level network logs. We propose a semi-supervised approach that combines hidden Markov models (HMM) with standard classifiers such as naive Bayes and support vector machines (SVM). This two step approach simplifies the occupancy model while achieving good accuracy. The experimental results over ten office cubicles show that the maximum error is less than 15% with an average error of 9.3%. We demonstrate that using our occupancy models, we can potentially reduce the lighting load on one floor (about 45 kW) by about 9.5%.

INTEGRATED NEURAL NETWORK MODELING FOR ELECTRONIC MANUFACTURING

This paper addresses issues involved in the modeling of electronic manufacturing processes for optimization and control using artificial neural networks (ANNs). A modeling methodology is presented which integrates a number of techniques to counter the commonly experienced problems of selecting the ‘right’ network structure, over-training and long training times in building economical and accurate ANN models. This methodology has been implemented as an automated user-friendly ANN modeling software — CU-ANN. The main features of our methodology are data pre-processing, ‘simple to complex’ network structure approach and simultaneous training and testing. The neural networks considered have feed forward architecture and use error back-propagation algorithm for training. We have successfully applied this ANN modeling methodology to a number of simulated and real-life electronic manufacturing problems. These include stencil printing and simulated wafer fab. process data. The results indicate that our approach produces accurate, economical models and can handle a wide variety of data sets.