Presentation is loading. Please wait.

Presentation is loading. Please wait.

GreenSoftware: Managing Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Md E Haque, Kien Le, Ryan Beauchea, Jordi Guitart, Jordi Torres,

Similar presentations


Presentation on theme: "GreenSoftware: Managing Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Md E Haque, Kien Le, Ryan Beauchea, Jordi Guitart, Jordi Torres,"— Presentation transcript:

1 GreenSoftware: Managing Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Md E Haque, Kien Le, Ryan Beauchea, Jordi Guitart, Jordi Torres, Thu D. Nguyen, Ricardo Bianchini Department of Computer Science

2 Motivation Datacenters consume large amounts of energy High energy cost and carbon footprint – Brown electricity: coal and natural gas Connect datacenters to green sources: solar, wind Apple DC in Maiden, NC 40MW solar farm 2

3 Challenges and opportunities Scheduling workload/energy sources – Lower costs: brown energy, peak brown power, capital Study opportunities in green datacenters – Build hardware/software Power Time Load Variable 3 Solar power Workload

4 GreenSoftware How to build software for green datacenters? 1.Malleable energy demand – Idle nodes → Turn off/Sleep (S3) [COLP’01] – Reduce frequency (DVFS) → Lower quality 2.Move computation under renewables – Weather forecast → Green energy forecast – Delay computation or degrade quality – Leverage energy storage 4

5 Outline Motivation GreenSoftware – GreenSlot – GreenHadoop – GreenSwitch – GreenCassandra – … and others Conclusion 5

6 GreenSlot [SC’11] Batch jobs on SLURM (& Hadoop) Send idle nodes to S3 Predict solar availability Delay jobs within deadlines – Known jobs characteristics (length, deadline, size…) – Heuristic 6 Time Job 1 Power Deadline Job 2 Job 3 Job 4

7 GreenSlot [SC’11] Batch jobs on SLURM (& Hadoop) Send idle nodes to S3 Predict solar availability Delay jobs within deadlines – Known jobs characteristics (length, deadline, size…) – Heuristic 7 Time Job 1 Power Deadline Job 2 Job 3 Job 4

8 GreenHadoop [Eurosys’12] Batch jobs on Hadoop Send idle nodes to S3 Make required data available – Move data blocks Predict solar availability Delay jobs within deadlines – Predict global jobs energy consumption – Heuristic 8 Map 1 2 3 4 5 Reduce 6 7 Shuffle

9 Covering subset GreenHadoop: Data management Deactivate servers to save energy – Some data might become unavailable Prior solution: covering subset [Leverich’09] – Set of servers always running has ALL data 9 7 3 45 216 8 7 1 45 6 3 2 8 1 7 3 Our approach Only required data has to be available We usually require fewer active servers Server Block

10 GreenHadoop: Data management Server 1 1 7 2 Active Decommission Down Server 2 4 35 6 Server 3 4 6 Required file Non-required file Server 4 2 3 8 4 Server 5 36 7 JobA 4 JobB 5 JobC 1 6 Running queue: 10

11 GreenHadoop: Data management Server 4 2 3 8 4 Server 5 36 7 Active Decommission Down GreenHadoop (computation) requires only 2 servers Server 1 1 7 2 1 7 2 Server 2 4 35 6 Server 3 4 6 Required file Non-required file JobA 4 JobB 5 JobC 1 6 Running queue: 11

12 GreenHadoop: Data management Active Decommission Down Move required files to Active servers Server 1 1 7 2 Server 2 4 35 6 Server 3 4 6 1 Server 4 2 3 8 4 Server 5 36 7 Replicate JobA 4 JobB 5 JobC 1 6 Running queue: 12

13 Server 1 1 7 2 GreenHadoop: Data management Active Decommission Down Decommissioned server can be sent to Down Server 1 1 7 2 Server 2 4 35 6 Server 3 4 6 Required file Non-required file 1 Server 4 2 3 8 4 Server 5 36 7 JobA 4 JobB 5 JobC 1 6 Running queue: 13

14 Server 1 1 7 2 GreenHadoop: Data management Active Decommission Down Jobs to be executed change → Required files change Server 2 4 35 6 Server 3 4 6 Non-required file 1 Server 4 2 3 8 4 Server 5 36 7 JobA 4 JobB 5 JobC 1 6 JobD 8 Required file 6 4 6 4 6 4 8 Running queue: 14

15 Server 4 2 3 8 4 Server 1 1 7 2 GreenHadoop: Data management Active Decommission Down Make missing data available Server 2 4 35 6 Server 4 2 3 8 4 Server 5 36 7 Server 3 4 6 1 Required file Non-required file JobB 5 JobC 1 JobD 8 Required file Running queue: 15

16 Server 4 2 3 8 4 Server 1 1 7 2 GreenHadoop: Data management Active Decommission Down Server 2 4 35 6 Server 4 2 3 8 4 Server 5 36 7 GreenHadoop (computation) requires 3 servers Server 3 4 6 1 Non-required file JobB 5 JobC 1 JobD 8 Required file Running queue: 16

17 GreenSwitch [ASPLOS’13] Batch jobs on Hadoop Similar to GreenHadoop Energy storage – Battery – Net metering Schedule workload and energy sources – Optimization Evaluation on Parasol (Presented on Monday by Thu) 17

18 GreenCassandra Distributed DB/storage on Cassandra Add an optional ring Degrade quality when no green 18 1 4 6 2 3 5 DHT Ring A A A 1 4 4 3 5 6 6 2 2 Double DHT Ring Optional A A Server Data

19 GreenSoftware summary TypeMalleable energyGreen adaptability GreenSlotBatch jobs Delay jobs Sleep servers Delay until green GreenHadoopBatch jobs Delay jobs Sleep servers Data management Delay until green GreenSwitchBatch/interactive jobs Delay jobs Sleep servers Delay until green Energy storage GreenCassandraDistributed storageOptional ringDegrade quality GreenSLAVMs Migrate VMs Sleep servers Route green energy to racks GreenParMPI jobs Change parallelism Sleep servers Greater parallelism on green GreenScaleNon-deferrable jobsCPU and mem DVFSFaster on green GreenNebulaGeo distributed VMsMigrate VMs“Follow the renewables” 19

20 Conclusions Green datacenters – Challenges & opportunities – Hardware/software solution GreenSoftware – Adapt software to green datacenters – Malleable energy demand – Match computation and renewables 20

21 GreenSoftware: Managing Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Md E Haque, Kien Le, Ryan Beauchea, Jordi Guitart, Jordi Torres, Thu D. Nguyen, Ricardo Bianchini Department of Computer Science

22 Other GreenSoftware GreenSLA [IGCC’13] – Bringing green energy to users – New hardware to route green energy GreenPar – MPI jobs with sub linear speedup – Use “Free” green energy GreenNebula – VMs in multiple geo distributed datacenters – Follow the sun GreenScale – Change frequency (DVFS) 22

23 GreenPar MPI jobs on VMs Add more computation capacity – Lower energy-efficiency – Use available “green” 23 Computation capacity Speedup Power Time

24 GreenSLA [IGCC’13] HPC jobs on VMs Users require % of green energy – Hardware to bring green energy to VMs Scheduling – Assign green energy to racks – Move VMs among racks – Heuristic 24

25 Default Green Datacenter Solar Inverter PDU Rack3 PDU Rack2 PDU Rack1 Mixed Bus Datacenter Power Distribution

26 GreenSLA: Proposed Power Distribution PDU Rack3 PDU Rack2 PDU Rack1 Battery Mixed Bus Control Module Charge Controller Green Bus SSS Solar Inverter Datacenter Power Distribution

27 Parasol without GreenSwitch Green use Green available Net metering Brown use IT load 27

28 GreenSwitch: deferrable workload Battery discharge Battery charge IT load Net metering 28 Green available Green use


Download ppt "GreenSoftware: Managing Datacenters Powered by Renewable Energy Íñigo Goiri, William Katsak, Md E Haque, Kien Le, Ryan Beauchea, Jordi Guitart, Jordi Torres,"

Similar presentations


Ads by Google