Mastering system and power measures for servers in datacenter

Costa, Georges Da; Pierson, Jean‐Marc; Cupertino, Leandro Fontoura

doi:10.1016/j.suscom.2017.05.003

Cited by 7 publications

(11 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Data Preprocessingmentioning

confidence: 99%

“…psu_model This method uses the PSU conversion losses model proposed in Reference [15] to filter the acquired data. timesync This method removes the timing jitters by the insertion of a synchronization pattern before the execution of each workload as described in Reference [15]. unique This method removes sequentially repeated values, keeping the first value and removing subsequent repeated values, as they come from the fact that watt-meters usually give outdated values if they are requested too frequently [15].…”

Section: Data Preprocessingmentioning

confidence: 99%

“…timesync This method removes the timing jitters by the insertion of a synchronization pattern before the execution of each workload as described in Reference [15]. unique This method removes sequentially repeated values, keeping the first value and removing subsequent repeated values, as they come from the fact that watt-meters usually give outdated values if they are requested too frequently [15]. steady The impact of the previously identified issues can be circumvented by the detection of steady states.…”

Section: Data Preprocessingmentioning

confidence: 99%

See 2 more Smart Citations

Effectiveness of Neural Networks for Power Modeling for Cloud and HPC

Costa

Pierson

Fontoura‐Cupertino

2020

ACM Trans. Model. Perform. Eval. Comput. Syst.

Self Cite

View full text Add to dashboard Cite

Power consumption of servers and applications are of utmost importance as computers are becoming ubiquitous, from smart phones to IoT and full-fledged computers. To optimize their power consumption, knowledge is necessary during execution at different levels: for the Operating System to take decisions of scheduling, for users to choose between different applications. Several models exist to evaluate the power consumption of computers without relying on actual wattmeters: Indeed, these hardware are costly but also usually have limits on their pooling frequency (usually a one-second frequency is observed) except for dedicated professional hardware. The models link applications behavior with their power consumption, but up to now there is a 5% wall: Most models cannot reduce their error under this threshold and are usually linked to a particular hardware configuration. This article demonstrates how to break the 5% wall of power models. It shows that by using neural networks it is possible to create models with 1% to 2% error. It also quantifies the reachable precision obtainable with other classical methods such as analytical models.

show abstract

Section: Data Preprocessingmentioning

confidence: 99%

Section: Data Preprocessingmentioning

confidence: 99%

Section: Data Preprocessingmentioning

confidence: 99%

See 1 more Smart Citation

Effectiveness of Neural Networks for Power Modeling for Cloud and HPC

Costa

Pierson

Fontoura‐Cupertino

2020

ACM Trans. Model. Perform. Eval. Comput. Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Measuring power consumption is done using external hardware and must be precisely correlated with system monitoring and must take into account several bias (such as time jitter, imprecision of hardware measurement, and so on). We followed a measurement methodology 11 to cope with this bias. The frequency of measurement is 1 second, sufficient for linking power consumption to behavior of HPC applications usually running from several minutes to several hours.…”

Section: Base Systemmentioning

confidence: 99%

Fast maximum coverage of system behavior from a performance and power point of view

Costa

Pierson

Fontoura‐Cupertino

2020

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

This article proposes a methodology for constructing and assessing the quality of benchmark-suites dedicated to understand server behavior. Applications are multiple: behavioral models creation, power capping, abnormal behavior detection, model validation. To reach all the operating points of one server, an automatic operating range detection is necessary. To get the knowledge of the operating range, an exhaustive search of possible reachable states must be conducted. While many works use for this purpose only simplistic benchmark-suites in their research, we show in this article that using simple assumptions (leading to simple models such as linear ones) introduces large bias. We identify the bias due to individual hardware components. The key for understanding and modeling of the system behavior (from performance and power consumption points of view) is to stress the system and its subsystems on a large set of configurations, and to collect values spanning over a large spectrum. We define a coverage metric for evaluating the coverage of measured performance indicators and power values. We evaluate different benchmarks using that metric, and thoroughly analyze their impact on the collected values. Finally, we propose a benchmark-suite providing a large coverage, adapted to general cases.

show abstract

“…Depending on the measurement hardware and infrastructure, biases such as the overhead of the monitoring software, PSU conversion impact, among other issues, have to be taken into account using, for instance, the methodology described in previous work [19].…”

Section: External Devicesmentioning

confidence: 99%

Energy monitoring as an essential building block towards sustainable ultrascale systems

Almeida

Assunção

Barbosa

et al. 2018

Sustainable Computing: Informatics and Systems

Self Cite

View full text Add to dashboard Cite

An Ultrascale System (USS) joins parallel and distributed computing systems that will be two to three orders of magnitude larger than today's infrastructure regarding scale, performance, the number of components and their complexity. For such systems to become a reality, however, advances must be made in High Performance Computing (HPC), largescale distributed systems, and big data solutions, also tackling challenges such as improving the energy efficiency of the IT infrastructure. Monitoring the power consumed by underlying IT resources is essential towards optimising the manner IT resources are used and hence improve the sustainability of such systems. Nevertheless, monitoring the energy consumed by USSs is a challenging endeavour as the system can comprise thousands of heterogeneous server resources spanning multiple data centres. Moreover, the amount of monitoring data, its gathering, and processing, should never become a bottleneck nor profoundly impact the energy efficiency of the overall system. This work surveys state of the art on energy monitoring of large-scale systems and methodologies for monitoring the power consumed by large systems and discusses some of the challenges to be addressed towards monitoring and improving the energy efficiency of USSs. Next, we present efforts made on designing monitoring solutions $ . Finally, we discuss potential gaps in existing solutions when tackling emerging large-scale monitoring scenarios and present some directions for future research on the topic.

show abstract

Mastering system and power measures for servers in datacenter

Cited by 7 publications

References 13 publications

Effectiveness of Neural Networks for Power Modeling for Cloud and HPC

Effectiveness of Neural Networks for Power Modeling for Cloud and HPC

Fast maximum coverage of system behavior from a performance and power point of view

Energy monitoring as an essential building block towards sustainable ultrascale systems

Contact Info

Product

Resources

About