High Performance Computing Data Centers Research Articles

Implementing the data center energy productivity metric

As data centers proliferate in size and number, the endeavor to improve their energy efficiency and productivity is becoming increasingly important. We discuss the properties of a number of the proposed metrics of energy efficiency and productivity. In particular, we focus on the Data Center Energy Productivity (DCeP) metric, which is the ratio of useful work produced by the data center to the energy consumed performing that work. We describe our approach for using DCeP as the principal outcome of a designed experiment using a highly instrumented, high-performance computing data center. We found that DCeP was successful in clearly distinguishing different operational states in the data center, thereby validating its utility as a metric for identifying configurations of hardware and software that would improve (or even maximize) energy productivity. We also discuss some of the challenges and benefits associated with implementing the DCeP metric, and we examine the efficacy of the metric in making comparisons within a data center and among data centers.

Rack aware scheduling in HPC data centers: an energy conservation strategy

Energy consumption in high performance computing data centers has become a long standing issue. With rising costs of operating the data center, various techniques need to be employed to reduce the overall energy consumption. Currently, among others there are techniques that guarantee reduced energy consumption by powering on/off the idle nodes. However, most of them do not consider the energy consumed by other components in a rack. Our study addresses this aspect of the data center. We show that we can gain considerable energy savings by reducing the energy consumed by these rack components. In this regard, we propose a scheduling technique that will help schedule jobs with the above mentioned goal. We claim that by our scheduling technique we can reduce the energy consumption considerably without affecting other performance metrics of a job. We implement this technique as an enhancement to the well-known Maui scheduler and present our results. We propose three different algorithms as part of this technique. The algorithms evaluate the various trade-offs that could be possibly made with respect to overall cluster performance. We compare our technique with various currently available Maui scheduler configurations. We simulate a wide variety of workloads from real cluster deployments using the simulation mode of Maui. Our results consistently show about 7 to 14 % savings over the currently available Maui scheduler configurations. We shall also see that our technique can be applied in tandem with most of the existing energy aware scheduling techniques to achieve enhanced energy savings. We also consider the side effects of power losses due to the network switches as a result of deploying our technique. We compare our technique with the existing techniques in terms of the power losses due to these switches based on the results in Sharma and Ranganathan, Lecture Notes in Computer Science, vol. 5550, 2009 and account for the power losses. We there on provide a best fit scheme with the rack considerations. We then propose an enhanced technique that merges the two extremes of node allocation based on rack information. We see that we can provide a way to configure the scheduler based on the kind of workload that it schedules and reduce the effect of job splitting across multiple racks. We further discuss how the enhancement can be utilized to build a learning model which can be used to adaptively adjust the scheduling parameters based on the workload experienced.

Energy-aware job scheduler for high-performance computing

In recent years energy-aware computing has become a major topic, not only in wireless and mobile devices but also in devices using wired technology. The ICT industry is consuming an increasing amount of energy and a large part of the consumption is generated by large-scale data centers. In High-Performance Computing (HPC) data centers, higher performance equals higher energy consumption. This has created incentives on exploring several alternatives to reduce the energy consumption of the system, such as energy-efficient hardware or the Dynamic Voltage and Frequency Scaling (DVFS) technique. This work presents an energy-aware scheduler that can be applied to a HPC data center without any changes in hardware. The scheduler is evaluated with a simulation model and a real-world HPC testbed. Our experiments indicate that the scheduler is able to reduce the energy consumption by 6---16% depending on the job workload. More importantly, there is no significant slowdown in the turnaround time or increase in the wait time of the job. The results hereby evidence that our approach can be beneficial for HPC data center operators without a large penalty on service level agreements.

Integrating cooling awareness with thermal aware workload placement for HPC data centers

High Performance Computing (HPC) data centers are becoming increasingly dense; the associated power-density and energy consumption of their operation are increasing. Up to half of the total energy is attributed to cooling the data center; greening the data center operations to reduce both computing and cooling energy is imperative. To this effect, this paper integrates awareness of the dynamic behavior of the cooling unit with thermal awareness while performing spatial workload scheduling (i.e. workload placement) in HPC data centers. The paper first proposes a coordinated cooling-aware job placement and cooling management algorithm, Highest Thermostat Setting (HTS). HTS is aware of dynamic behavior of the Computer Room Air Conditioner (CRAC) units and places jobs to reduce cooling demands from the CRACs. HTS also dynamically updates the CRAC thermostat set point to reduce cooling energy consumption. Further, the Energy Inefficiency Ratio of SPatial job scheduling (a.k.a. job placement) algorithms, also referred as SP-EIR, is analyzed by comparing the total (computing + cooling) energy consumption incurred by the algorithms with the minimum possible energy consumption, while assuming that the job start times are already decided to meet the Service Level Agreements (SLAs). This analysis is performed for two cooling models, constant and dynamic, to show how the constant cooling model assumption in previous research misses out on opportunities to save energy. Simulation results based on power measurements and job traces from the ASU HPC data center show that: (i) HTS has 15% lower SP-EIR compared to LRH, a thermal-aware spatial scheduling algorithm; and (ii) in conjunction with FCFS-Backfill, HTS increases the throughput per unit energy by 6.89% and 5.56%, respectively, over LRH and MTDP (an energy-efficient spatial scheduling algorithm with server consolidation).

Energy-Efficient Thermal-Aware Task Scheduling for Homogeneous High-Performance Computing Data Centers: A Cyber-Physical Approach

High-performance computing data centers have been rapidly growing, both in number and size. Thermal management of data centers can address dominant problems associated with cooling such as the recirculation of hot air from the equipment outlets to their inlets and the appearance of hot spots. In this paper, we show through formalization that minimizing the peak inlet temperature allows for the lowest cooling power needs. Using a low-complexity linear heat recirculation model, we define the problem of minimizing the peak inlet temperature within a data center through task assignment (MPIT-TA), consequently leading to minimal cooling-requirement. We also provide two methods to solve the formulation: Xlnt-GA, which uses a genetic algorithm, and Xlnt-SQP, which uses sequential quadratic programming. Results from small-scale data center simulations show that solving the formulation leads to an inlet temperature distribution that, compared to other approaches, is 2 degC to 5 degC lower and achieves about 20 to 30 percent cooling energy savings at common data center utilization rates. Moreover, our algorithms consistently outperform the minimize heat recirculation algorithm, a recirculation-reducing task placement algorithm in the literature.