1 / 20

Green Computing

Green Computing. Metrics: Power, Temperature, CO2, … Computing system: Many-cores, Clusters, Grids and Clouds Algorithm and model: task scheduling, CFD model, … Middleware: auditing & insertion service, green resource management service, ….

mandar
Download Presentation

Green Computing

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Green Computing Metrics: Power, Temperature, CO2, … Computing system: Many-cores, Clusters, Grids and Clouds Algorithm and model: task scheduling, CFD model, … Middleware: auditing & insertion service, green resource management service, …

  2. Power aware virtual machine scheduling in a DVFS cluster Virtual machine in Grids and Clouds Dynamic Voltage Frequency Scheduling Objective: dynamically scale voltages for virtual machines in a cluster

  3. Virtual machines in compute cluster job Execute job in a vm Start a vm vm vm vm vm vm vm File server Head node Compute node Compute node

  4. Schedule virtual machines VM PE PE PE Scheduling algorithm PE queue PE PE cluster

  5. Power aware scheduling algorithm Sort VMs in a decreasing order of required CPU speed Set PEs to lowest voltages put VMs to PEs If cannot accommodate, level up PE voltages Level down PE voltages whenever it is possible to accommodate VMs

  6. Simulation Results

  7. Experimental Results

  8. Thermal aware workload scheduling in data centers Job-temperature model Data center resource model Thermal aware scheduling algorithm (TASA) Thermal aware workload scheduling framework Simulation

  9. Job-temperature profile

  10. Data center model (1) Z Y Node (x,y,z) Rack Rack Rack Hot air Hot air X

  11. Data center model (2)

  12. Thermal aware scheduling framework Profiling tool Workload model Task-temperature profile Workload placement Thermal aware workload scheduling algorithm Online task-temperature calculation Datacenter model Cooling system control RC-thermal model Thermal map CFD model Monitoring service

  13. Thermal aware scheduling algorithm (TASA) Get thermal field of data center Get compute node temperature Put hottest job to coldest resources Predict the compute node temperature after job execution If a compute node temperature > redline, set it idle thermal aware backfilling when it is possible

  14. Simulation Real workload logged in CCR @ Buffalo Univ. Temperature logged FCFS in CCR @ Buffalo Univ. TASA Discussion

  15. Workload in CCR

  16. CCR Temperature

  17. Simulation Result (1) Reduce max temperature: 6 F Reduce average temperature: 15 F Reduce power consumption 4000 kW/h Reduce CO2 emission 19 000 kg

  18. Simulation Result (2) Response time increase 13%

  19. Green Data Center Computing: concept CFD model Workload model Thermal aware resource management Auditing & Insertion service Auditing & Insertion service Software sensor Physical sensor Monitoring service Cooling system and compute resources in a data center

  20. Cyberaide Green: Software achitecture Client Layer Python Client Java CoG Kit Cyberaide Portal Secure Web Service Cyberaide Shell Command Line Authentication and Authority Middleware Layer Workflow Information collector Thermal-Aware Meta Scheduler Task Submission Information Secure Web Service Resource Layer Data Center Data Center

More Related