Energy Usage in Cloud Part 2 Salih Safa

Energy Usage in Cloud Part 2 Salih Safa BACANLI

• • Cooling Virtualization Energy Proportional System Conclusion

Cooling • After servers, the second largest consumer of power in a data center is the cooling system. (Kava, 2012) • This cloud-computing hardware must be cooled by something, regardless of the hardware’s location or owner. • Required cooling infrastructure depends on the computing resources’ power density and layout. (Infotech, 2011)

• Air cooling is popular, inexpensive, simple to scale to fit room needs. Fine for systems up to 15 KW • You need to move the air through the equipment (air’s heat transfer coefficient is nearly constant). – Ventilation air and natural winds OR • You need to cool the air in the cloud room.

Raised Floor • The equipments (racks) are put on a platform higher than the ground. The area is used for air space or water cooling pipes.

• For systems who energy consumption is bigger than 15 KW/rack you need to use liquid. Moving large volumes of air can be very expensive for very-high-density racks. • Mineral oil and water can be used for liquid.

Virtualization as solution? • Virtualization is nice if used properly. – Having many computers in the same hardware. • Scaling: – Horizontal: Increasing VM – Vertical: increase resource allocation of current VMs (increase their CPU share) (Paya & Marinescu, 2013)

Disadvantageous Virtualization 1) The rise of high density– Higher power density is likely to result from virtualization, at least in some racks. Areas of high density can pose cooling challenges that, if left unaddressed, could threaten the reliability of the overall data center. 2) Reduced IT load can affect PUE– After virtualization, the data center’s power usage effectiveness (PUE) is likely to worsen. This is despite the fact that the initial physical server consolidation results in lower overall energy use. If the power and cooling infrastructure is not right-sized to the new lower overall load, physical infrastructure efficiency measured as PUE will degrade.

3) Dynamic IT loads– Virtualized IT loads, particularly in a highly virtualized, cloud data center, can vary in both time AND location. In order to ensure availability in such a system, it’s critical that rack-level power and cooling health be considered before changes are made. 4)Lower redundancy requirements are possible– A highly virtualized data center designed and operated with a high level of IT fault-tolerance may reduce the necessity for redundancy in the physical infrastructure. This effect could have a significantly positive impact on data center planning and capital costs.

• To improve post-virtualization PUE, the data center’s infrastructure efficiency curve must be improved (lowered) by optimizing power and cooling systems to reduce the waste of oversizing and better align capacity with the new, lower load. In addition to improving efficiency, optimized power and cooling will directly impact the electric bill by reducing the power consumed by unused power and cooling capacity.

• Energy proportional system • Normal System

For normal server Utilization is measure of application perfomance like requests per second on a web server

Energy efficiency of servers have increased as the time passed.

For Energy Proportional Server

Conclusion • The introduction of more efficient CPUs based on chip multiprocessing has also contributed positively toward more energy-efficient servers. • However, long-term technology trends invariably indicate that higher performance means increased energy usage. As a result, energy efficiency must improve as fast as computing performance to avoid a significant growth in computers’ energy footprint. (Barroso & Hölzle, 2007)

To a first-order approximation, both cooling and provisioning costs are proportional to the average energy that servers consume, therefore energy efficiency improvements should benefit all energy dependent components

References N. A (2011). Cooling the Cloud, Infotech. [ONLINE] Retrived from http: //it. tmcnet. com/topics/it/articles/216010 -cooling-cloud. htm [24. 10. 2013] Kava, J. (2012). Cooling the cloud: A look inside Google’s Hot Huts. Google Green Blog[ONLINE]Retrieved from http: //googlegreenblogspot. com/2012/10/coolingcloud-look-inside-googles-hot. html [24. 10. 2013] Barroso, L. A. & Hölzle, U. (2007). The case for energy proportional computing. Computer, December 2007, 33 -37. Schneider Electricity-Data Center Science [ONLINE] Retrieved from http: //www. facilitiesnet. com/microsite/energy-efficient-data-centerstrategies/pdf/White. Paper 118. pdf [24. 10. 2013] Paya, A. & Marinescu, D. C. (2013). Energy-aware Application Scaling on a Cloud. Arxiv. org[ONLINE] Retrieved from http: //arxiv. org/abs/1307. 3306 [24. 10. 2013]