Workload placement on servers has been traditionally driven by mainly performance objectives. In this work, we investigate the design, implementation, and evaluation of a power-aware application placement controller in the context of an environment with heterogeneous virtualized server clusters. The placement component of the application management middleware takes into account the power and migration costs in addition to the performance benefit while placing the application containers on the physical servers. The contribution of this work is two-fold: first, we present multiple ways to capture the cost-aware application placement problem that may be applied to various settings. For each formulation, we provide details on the kind of information required to solve the problems, the model assumptions, and the practicality of the assumptions on real servers. In the second part of our study, we present the pMapper architecture and placement algorithms to solve one practical formulation of the problem: minimizing power subject to a fixed performance requirement. We present comprehensive theoretical and experimental evidence to establish the efficacy of pMapper.

/pdf/pmapper-power-and-migration-cost-aware-application-placement-2o53nt8erw.pdf

pMapper: power and migration cost aware application placement in virtualized systems

Energy consumption in hosting Internet services is becoming a pressing issue as these services scale up Dynamic server provisioning techniques are effective in turning off unnecessary servers to save energy Such techniques, mostly studied for request-response services, face challenges in the context of connection servers that host a large number of long-lived TCP connections In this paper, we characterize unique properties, performance, and power models of connection servers, based on a real data trace collected from the deployed Windows Live Messenger Using the models, we design server provisioning and load dispatching algorithms and study subtle interactions between them We show that our algorithms can save a significant amount of energy without sacrificing user experiences

/pdf/energy-aware-server-provisioning-and-load-dispatching-for-5edt7m0bxi.pdf

Energy-aware server provisioning and load dispatching for connection-intensive internet services

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.

http://www.cse.psu.edu/~axs53/csl/papers/metrics05a.pdf

Managing server energy and operational costs in hosting centers

Energy conservation without performance degradation is an important goal for battery-operated computers, such as laptops and hand-held assistants. In this paper we determine the potential benefits of application-supported device management for optimizing energy and performance. In particular, we consider application transformations that increase device idle times and inform the operating system about the length of each upcoming period of idleness. We use modeling and experimentation to assess the potential energy and performance benefits of this type of application support for a laptop disk. Our main modeling results show that these benefits are significant. Our experimental results demonstrate that unless applications are transformed, they cannot accrue any of the predicted benefits. Overall, we find that the transformations can reduce disk energy consumption by as much as 89% with only a small degradation in performance.

Energy-efficient server clusters

Server farms today consume more than 1.5% of the total electricity in the U.S. at a cost of nearly $4.5 billion. Given the rising cost of energy, many industries are now seeking solutions for how to best make use of their available power. An important question which arises in this context is how to distribute available power among servers in a server farm so as to get maximum performance.By giving more power to a server, one can get higher server frequency (speed). Hence it is commonly believed that, for a given power budget, performance can be maximized by operating servers at their highest power levels. However, it is also conceivable that one might prefer to run servers at their lowest power levels, which allows more servers to be turned on for a given power budget. To fully understand the effect of power allocation on performance in a server farm with a fixed power budget, we introduce a queueing theoretic model, which allows us to predict the optimal power allocation in a variety of scenarios. Results are verified via extensive experiments on an IBM BladeCenter.We find that the optimal power allocation varies for different scenarios. In particular, it is not always optimal to run servers at their maximum power levels. There are scenarios where it might be optimal to run servers at their lowest power levels or at some intermediate power levels. Our analysis shows that the optimal power allocation is non-obvious and depends on many factors such as the power-to-frequency relationship in the processors, the arrival rate of jobs, the maximum server frequency, the lowest attainable server frequency and the server farm configuration. Furthermore, our theoretical model allows us to explore more general settings than we can implement, including arbitrarily large server farms and different power-to-frequency curves. Importantly, we show that the optimal power allocation can significantly improve server farm performance, by a factor of typically 1.4 and as much as a factor of 5 in some cases.

/pdf/optimal-power-allocation-in-server-farms-9mb4fzr2li.pdf

Optimal power allocation in server farms

Energy management for servers is now necessary for technical, financial, and environmental reasons. This paper describes three policies designed to reduce energy consumption in Web servers. The policies employ two power management mechanisms: dynamic voltage scaling (DVS), an existing mechanism, and request batching, a new mechanism introduced in this paper. The first policy uses DVS in isolation, except that we extend recently introduced task-based DVS policies for use in server environments with many concurrent tasks. The second policy uses request batching to conserve energy during periods of low workload intensity. The third policy uses both DVS and request batching mechanisms to reduce processor energy usage over a wide range of workload intensities. All the policies trade off system responsiveness to save energy. However, the policies employ the mechanisms in a feedback-driven control framework in order to conserve energy while maintaining a given quality of service level, as defined by a percentile-level response time.

We evaluate the policies using Salsa, a web server simulator that has been extensively validated for both energy and response time against measurements from a commodity web server. Three daylong static web workloads from real web server systems are used to quantify the energy savings: the Nagano Olympics98 web server, a financial services company web site, and a disk intensive web workload. Our results show that when required to maintain a 90th-percentile response time of 50ms, the DVS and request batching policies save from 8.7% to 38% and from 3.1% to 27% respectively of the CPU energy used by the base system. The two polices provide these savings for complementary workload intensities. The combined policy is effective for all three workloads across a broad range of intensities, saving from 17% to 42% of the CPU energy.

Energy conservation policies for web servers

Mambo is a full-system simulator for modeling PowerPC-based systems. It provides building blocks for creating simulators that range from purely functional to timing-accurate. Functional versions support fast emulation of individual PowerPC instructions and the devices necessary for executing operating systems. Timing-accurate versions add the ability to account for device timing delays, and support the modeling of the PowerPC processor microarchitecture. We describe our experience in implementing the simulator and its uses within IBM to model future systems, support early software development, and design new system software.

Mambo: a full system simulator for the PowerPC architecture

The response time of a WWW service often plays an important role in its success or demise. From a user's perspective, the response time is the time elapsed from when a request is initiated at a client to the time that the response is fully loaded by the client. This paper presents a framework for accurately measuring the client-perceived response time in a WWW service. Our framework provides feedback to the service provider and eliminates the uncertainties that are common in existing methods. This feedback can be used to determine whether performance expectations are met, and whether additional resources (e.g. more powerful server or better network connection) are needed. The framework can also be used when a consolidator provides Web hosting service, in which case the framework provides quantitative measures to verify the consolidator's compliance to a specified Service Level Agreement. Our approach assumes the existing infrastructure of the Internet with its current technologies and protocols. No modification is necessary to existing browsers or servers, and we accommodate intermediate proxies that cache documents. The only requirement is to instrument the documents to be measured, which can be done automatically using a tool we provide.

Measuring client-perceived response times on the WWW

Fault Management Workshop Final Report, August 13, 2012

: The Power Aware Real-Time Scheduling (PARTS) project is based on attacking the power management problem for real-time systems at the system level. This includes both modifications to the applications and to the operating system as well as compiler modifications to insert so-called power management points. The PARTS project started with scheduling independent tasks in a single CPU and has expanded to multiple CPUs as well as tasks with dependencies. The authors approached the power management problem from three different angles. The first is the minimization of energy consumption for given timing constraints. The second is the maximization of the system's reward (utility) for specific timing and energy restrictions (including rechargeable systems). The third is the tradeoff between energy consumption and reliability for specific timing and performance restrictions. Further, they validated the theory developed by carrying out implementations on the Power Aware Multiprocessor Architecture platform (PAMA), which consists of four rad-hard PowerPC 750 processors running actual space applications.

Mootaz Elnozahy

Papers

Energy conservation policies for web servers

Mambo: a full system simulator for the PowerPC architecture

Measuring client-perceived response times on the WWW

Fault Management Workshop Final Report, August 13, 2012

Power Management for Real-Time Systems