Proceedings ArticleDOI
The case for sleep states in servers
Anshul Gandhi,Mor Harchol-Balter,Michael Kozuch +2 more
- pp 2
Reads0
Chats0
TLDR
The goal of this research is to encourage data center administrators to consider dynamic power management and to spur chip designers to develop useful sleep states for servers.Abstract:
While sleep states have existed for mobile devices and workstations for some time, these sleep states have largely not been incorporated into the servers in today's data centers.Chip designers have been unmotivated to design sleep states because data center administrators haven't expressed any desire to have them. High setup times make administrators fearful of any form of dynamic power management, whereby servers are suspended or shut down when load drops. This general reluctance has stalled research into whether there might be some feasible sleep state (with sufficiently low setup overhead and/or sufficiently low power) that would actually be beneficial in data centers.This paper uses both experimentation and theory to investigate the regime of sleep states that should be advantageous in data centers. Implementation experiments involve a 24-server multi-tier testbed, serving a web site of the type seen in Facebook or Amazon with key-value workload and a range of hypothetical sleep states. Analytical modeling is used to understand the effect of scaling up to larger data centers. The goal of this research is to encourage data center administrators to consider dynamic power management and to spur chip designers to develop useful sleep states for servers.read more
Citations
More filters
Journal ArticleDOI
AutoScale: Dynamic, Robust Capacity Management for Multi-Tier Data Centers
TL;DR: A dynamic capacity management policy, AutoScale, is introduced that greatly reduces the number of servers needed in data centers driven by unpredictable, time-varying load, while meeting response time SLAs and robustness.
Proceedings ArticleDOI
Are sleep states effective in data centers
TL;DR: The implementation results on a 24-server multi-tier testbed indicate that under many traces, sleep states greatly enhance dynamic power management, and simulation results suggest that sleep states are even more beneficial for larger data centers.
Proceedings ArticleDOI
Navigating heterogeneous processors with market mechanisms
TL;DR: This work designs and evaluates twelve design points along the Xeon-to-Atom spectrum, and finds that a mix of three processor architectures achieves a 12× reduction in response time violations relative to equal-power homogeneous systems.
Proceedings ArticleDOI
Energy-aware scheduling for infrastructure clouds
Thomas Knauth,Christof Fetzer +1 more
TL;DR: This work uses simulation to quantify the difference in energy consumption caused exclusively by virtual machine schedulers, and presents its own optimized scheduler, OptSched, which reduces cumulative machine uptime by up to 60.1%.
Proceedings ArticleDOI
Agile, efficient virtualization power management with low-latency server power states
Canturk Isci,Suzanne McIntosh,Jeffrey O. Kephart,Rajarshi Das,James E. Hanson,Scott A. Piper,Robert R. Wolford,Thomas M. Brey,Robert F. Kantner,Allen Ng,James Norris,Abdoulaye Traore,Michael J. Frissora +12 more
TL;DR: It is presented, via both real system implementations and scale-out simulations, that virtualization power management with low-latency server power states can achieve comparable overheads as base distributed resource management in virtualized systems, and thus can benefit from the same level of adoption, while delivering close to energy-proportional power efficiency.
References
More filters
Journal ArticleDOI
The Case for Energy-Proportional Computing
Luiz Andre Barroso,Urs Hölzle +1 more
TL;DR: Energy-proportional designs would enable large energy savings in servers, potentially doubling their efficiency in real-life use, particularly the memory and disk subsystems.
Proceedings ArticleDOI
PowerNap: eliminating server idle power
TL;DR: The PowerNap concept, an energy-conservation approach where the entire system transitions rapidly between a high-performance active state and a near-zero-power idle state in response to instantaneous load, is proposed and the Redundant Array for Inexpensive Load Sharing (RAILS) is introduced.
Journal ArticleDOI
httperf—a tool for measuring web server performance
David Mosberger,Tai Jin +1 more
TL;DR: In this article, the authors describe a tool for measuring web server performance called httperf, which provides a flexible facility for generating various HTTP workloads and for measuring server performance.
Proceedings ArticleDOI
Power management of online data-intensive services
David Meisner,Christopher M. Sadler,Luiz Andre Barroso,Wolf-Dietrich Weber,Thomas F. Wenisch +4 more
TL;DR: This work evaluates the applicability of active and idle low-power modes to reduce the power consumed by the primary server components (processor, memory, and disk), while maintaining tight response time constraints, particularly on 95th-percentile latency.
Proceedings Article
Berkeley DB
TL;DR: The design and technical features of Berkeley DB, the distribution, and its license are described, including surviving system and disk crashes.