Natarajan Gautam

Managing server energy and operational costs in hosting centers

The growing cost of tuning and managing computer systems is leading to out-sourcing of commercial services to hosting centers. These centers provision thousands of dense servers within a relatively small real-estate in order to host the applications/services of different customers who may have been assured by a service-level agreement (SLA). Power consumption of these servers is becoming a serious concern in the design and operation of the hosting centers. The effects of high power consumption manifest not only in the costs spent in designing effective cooling systems to ward off the generated heat, but in the cost of electricity consumption itself. It is crucial to deploy power management strategies in these hosting centers to lower these costs towards enhancing profitability. At the same time, techniques for power management that include shutting down these servers and/or modulating their operational speed, can impact the ability of the hosting center to meet SLAs. In addition, repeated on-off cycles can increase the wear-and-tear of server components, incurring costs for their procurement and replacement. This paper presents a formalism to this problem, and proposes three new online solution strategies based on steady state queuing analysis, feedback control theory, and a hybrid mechanism borrowing ideas from these two. Using real web server traces, we show that these solutions are more adaptive to workload behavior when performing server provisioning and speed control than earlier heuristics towards minimizing operational costs while meeting the SLAs.

A Three-Dimensional Computer Simulation Model Reveals the Mechanisms for Self-Organization of Plant Cortical Microtubules into Oblique Arrays

We use a 3D computer simulation model that is based on experimental data to understand how the noncentrosomal plant cortical microtubules self-organize into specific ordered patterns in both wild-type and mutant plants.

Synthesizing Representative I/O Workloads for TPC-H

Synthesizing I/O requests that can accurately capture workload behavior is extremely valuable for the design, implementation and optimization of disk subsystems. This paper presents a synthetic workload generator for TPC-H, an important decision-support commercial workload, by completely characterizing the arrival and access patterns of its queries. We present a novel approach for parameterizing the behavior of inter-mingling streams of sequential requests, and exploit correlations between multiple attributes of these requests, to generate disk block-level traces that are shown to accurately mimic the behavior of a real trace in terms of response time characteristics for each TPC-H query.

On Queues with Markov Modulated Service Rates

In this paper, we consider a queue whose service speed changes according to an external environment that is governed by a Markov process. It is possible that the server changes its service speed many times while serving a customer. We derive first and second moments of the service time of customers in system using first step analysis to obtain an insight on the service process. In fact, we obtain an intriguing result in that the moments of service time actually depend on the arrival process! We also show that the mean service rate is not the reciprocal of the mean service time.Further, since it is not possible to obtain a closed form expression for the queue length distribution, we use matrix geometric methods to compute performance measures such as average queue length and waiting time. We apply the method of large deviations to obtain tail distributions of the workload in the queue using the concept of effective bandwidth. We present two applications in computer systems: (1) Web server with multi-class requests and (2) CPU with multiple processes. We illustrate the analysis and various methods discussed with the help of numerical examples for the above two applications.

Analysis of Queues: Methods and Applications

Written with students and professors in mind, Analysis of Queues: Methods and Applications combines coverage of classical queueing theory with recent advances in studying stochastic networks. Exploring a broad range of applications, the book contains plenty of solved problems, exercises, case studies, paradoxes, and numerical examples. In addition to the standard single-station and single class discrete queues, the book discusses models for multi-class queues and queueing networks as well as methods based on fluid scaling, stochastic fluid flows, continuous parameter Markov processes, and quasi-birth-and-death processes, to name a few. It describes a variety of applications including computer-communication networks, information systems, production operations, transportation, and service systems such as healthcare, call centers and restaurants.

Admission control of multi-class traffic with service priorities in high-speed networks

We consider a fluid model of a system that handles multiple classes of traffic. The delay and cell-loss requirements of the different classes of traffic are generally widely different and are achieved by assigning different buffers for different classes, and serving them in a strict priority order. We use results from the effective bandwidth of the output processes (see Chang and Thomas (1995)) to derive simple and asymptotically exact call-admission policies for such a system to guarantee the cell-loss requirements for the different classes assuming that each source produces a single class traffic. We compare the admission-control policies developed here with the approximate policy studied by Elwalid and Mitra (1995) for the case of two-class traffic.

Modeling and analysis of dynamic coscheduling in parallel and distributed environments

Scheduling in large-scale parallel systems has been and continues to be an important and challenging research problem. Several key factors, including the increasing use of off-the-shelf clusters of workstations to build such parallel systems, have resulted in the emergence of a new class of scheduling strategies, broadly referred to as dynamic coscheduling. Unfortunately, the size of both the design and performance spaces of these emerging scheduling strategies is quite large, due in part to the numerous dynamic interactions among the different components of the parallel computing environment as well as the wide range of applications and systems that can comprise the parallel environment. This in turn makes it difficult to fully explore the benefits and limitations of the various proposed dynamic coscheduling approaches for large-scale systems solely with the use of simulation and/or experimentation.To gain a better understanding of the fundamental properties of different dynamic coscheduling methods, we formulate a general mathematical model of this class of scheduling strategies within a unified framework that allows us to investigate a wide range of parallel environments. We derive a matrix-analytic analysis based on a stochastic decomposition and a fixed-point iteration. A large number of numerical experiments are performed in part to examine the accuracy of our approach. These numerical results are in excellent agreement with detailed simulation results. Our mathematical model and analysis is then used to explore several fundamental design and performance tradeoffs associated with the class of dynamic coscheduling policies across a broad spectrum of parallel computing environments.

Computer simulation and mathematical models of the noncentrosomal plant cortical microtubule cytoskeleton

There is rising interest in modeling the noncentrosomal cortical microtubule cytoskeleton of plant cells, particularly its organization into ordered arrays and the mechanisms that facilitate this organization. In this review, we discuss quantitative models of this highly complex and dynamic structure both at a cellular and molecular level. We report differences in methodologies and assumptions of different models as well as their controversial results. Our review provides insights for future studies to resolve these controversies, in addition to underlining the common results between various models. We also highlight the need to compare the results from simulation and mathematical models with quantitative data from biological experiments in order to test the validity of the models and to further improve them. It is our hope that this review will serve to provide guidelines for how to combine quantitative and experimental techniques to develop higher‐level models of the plant cytoskeleton in the future.

Opportunities for network coding: To wait or not to wait

It has been well established that reverse-carpooling based network coding can significantly improve the efficiency of multi-hop wireless networks. However, in a stochastic environment when there are no opportunities to code because of packets without coding pairs, should these packets wait for a future opportunity or should they be transmitted without coding? To help answer that question we formulate a stochastic dynamic program with the objective of minimizing the long-run average cost per unit time incurred due to transmissions and delays. In particular, we develop optimal control actions that would balance between costs of transmission against those of delays. In that process we seek to address a crucial question: what should be observed as the state of the system? We analytically show that just the queue lengths is enough if it can be modeled as a Markov process. Subsequently we show that a stationary policy based on queue lengths is optimal and describe a procedure to find such a policy. We further substantiate our results with simulation experiments for more generalized settings.

Combined Routing and Node Replacement in Energy-Efficient Underwater Sensor Networks for Seismic Monitoring

Ocean-bottom seismic systems are emerging as superior information-acquisition methods in seismic monitoring of petroleum reservoirs beneath ocean beds. These systems use a large network of sensor nodes that are laid on the ocean floor to collectively gather and transmit seismic information. In particular, underwater wireless sensor networks are gaining prominence in continuous seismic monitoring of undersea oilfields. They are autonomous and use wireless acoustic transmission for transferring data. However, the deployment period of such networks extends well beyond the battery lifetimes of the nodes. Hence, to ensure continuous monitoring from all node locations, it is required to replace the energy-depleted nodes on the ocean floor. Replacing these nodes at remote undersea locations is very expensive, and hence, the total node replacement cost in large seismic node networks is extremely high. In this paper, we develop effective joint policies involving routing and node replacement decisions to minimize the replacement costs per unit time. Our routing approach is simple and suitable for this application, and it strives for energy efficiency at all times. We propose a few node-replacement policies and compare their performances when they are combined with our proposed routings.