T : A Data-Centric, Cyber-Physical Cooling Energy Costs Reduction

Kaushik, Rini T.

Permalink

https://hdl.handle.net/2142/30802 Copy

Description

Title

T : A Data-Centric, Cyber-Physical Cooling Energy Costs Reduction

Author(s)

Kaushik, Rini T.

Contributor(s)

Nahrstedt, Klara

Issue Date

2011-05-01

Keyword(s)

• Big Data
• Energy
• Hadoop
• HDFS
• Big Data Analytics
• Cooling Energy Costs
• Cooling
• Thermal
• Data Placement
• File System

Abstract

Explosion in Big Data has led to a surge in extremely large-scale Big Data Analytic platforms, resulting in burgeoning energy costs. Big Data compute model mandates strong data-locality for computation performance due to Big Data inertia and bandwidth constraints, and moves computations to data. State-of-the-art cooling energy management techniques rely on thermal-aware computational job placement/migration and are inherently data agnostic in nature. T* takes a novel data-centric approach to reduce cooling energy costs. On the physical side, T* is cognizant of the thermal-profile in the data centers. On the cyber-side, T* is aware of the differences in the data-semantics (i.e., computational job profile, job rate, and life spans) of the Big Data placed in the compute cloud. Based on this knowledge, and coupled with its predictive data models, T* does proactive, cyber-physical, thermal-aware data placement, which implicitly results in thermal-aware job placement in the Big Data Analytics cloud compute model. Evaluation results with one-month long real-world Big Data analytics production traces show up to 59% reduction in the cooling energy costs with T* while performing 9x better than the state-of-the-art data agnostic cooling techniques in the Big Data environment.

Type of Resource

text

Language

en

Permalink

http://hdl.handle.net/2142/30802

Owning Collections