The document presents research on energy-efficient parallel storage systems utilizing write-buffer disks, aimed at reducing power consumption in data centers. It discusses the architectural design and simulation of a buffer disk system to optimize energy efficiency while managing performance. The findings illustrate the impact of workload on energy consumption and suggest strategies to maintain disks in standby mode to minimize energy usage.

1. Energy-Efficient Parallel Storage Systems with Write-Buffer DisksXiaojunRuan and Xiao QinComputer Science and Software EngineeringSamuel Ginn College of EngineeringAuburn University

2. My Research Group: 2011

3. Overview of the ProjectEnergy EfficiencySecuritySolid State DrivesPerformance of Secure Disk Systems[IEEE NAS09]Design, Model, Simulate, And Evaluate Disk Systems with Buffer Disks[ACM SAC09][ICPP09]Enhancing Internal Parallelism of SSDs[To Be Submitted11]Message Passing Interface with Enhanced Security[IPCCC 2010]Energy-EfficientDistributed StorageSystems [IPCCC10]Energy-Efficient Dynamic Voltage Scaling[ICCCN07] Xiaojun Ruan3BUD

4. 46/10/2011Electricity Usage in Data CentersAnnual Data Center Electricity Usage and Electricity Price increase Every year

5. The average power consumption of TOP10 supercomputing systems is 1.32 Mwatt.Dell’s Texas Data Center56/10/2011Energy Efficiency of Supercomputers

6. Electrical Cost of Data CentersUsing 2010 Trends Scenario

7. Server and Data Centers Consume 110 Billion kWh per year

8. Assume average commercial end user is charged 9.46 kWh

9. Disk systems can account for 27% of the energy cost of data centers6/10/20116Server and data centers may have an electrical cost of 10.4 billion dollars.

10. 76/10/2011Energy Consumption of Disks

11. Power States of Disks Active State: high energy consumptionActiveStandbyState transition penalty Standby State: low energy consumption8

12. A Hard Disk DriveA10000RPM Hard Drive may take 10.9 seconds to wake up!9

13. Parallel DisksPerformanceEnergy Efficiency

14. ChallangesPerformance Oriented: Best Performance

15. Huge Electricity Bills Energy Efficiency Oriented: Worst Performance?

16. Small Electricity BillsBasic Idea of BUDKeep Disks in Standby mode as long as possible

17. Reduce Status Transitions as many as possible12

18. IBM Ultrastar 36Z156/10/201113

19. A Parallel Disk System with a Write Buffer Disk

20. The BUD ArchitectureData Disks can serve requests without buffer disks when workload is high15

21. Auburn University 16Sum of Requests in Buffer (SRB)SRB is Number of the buffered requests targetingat the same data disk.

22. SRB is set by administrators

23. Once SRB is satisfied, spin up the targeted data disk, dump all those data, then spin the disk down.Scheduling StrategyTo buffer enough requests targeting at the same data diskDynAmic Request Allocation algorithm for Writes17

24. ExampleBuffer DiskRequests Queue18

25. Auburn University Xiaojun Ruan19From Design to Simulation

26. Simulation Environment20

27. Auburn University 21Workloads

28. Impact of SRB—Low Workload, UltraStar22

29. Auburn University Xiaojun Ruan23Non-Buffer Experiments

30. Auburn University 24BUD with IBM 40GNX TravalStar

31. Buffer Disk Number and Workload-- UltraStar25

32. Auburn University 26Energy ConsumptionE = Active Energy Consumption + Standby Energy Consumption + Transition Penalty

33. Auburn University Xiaojun Ruan27From Simulation to Real Implementation

34. An Energy-Efficient Cluster Storage System28

35. Implementation (no buffer disks)29

36. Implementation (with buffer-disks)30

37. Experimental DesignDisk Category I/O Node 1

38. Data Disk 1: WesternDigital 400, 20GB

39. Data Disk 2: WesternDigital 400, 20GB

40. Disk Category I/O Node 2