DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper...

44
SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales Ke Wang Data-Intensive Distributed Systems Laboratory Computer Science Department Illinois Institute of Technology February 14 th , 2012

Transcript of DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper...

Page 1: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

SimMatrix: SIMulator for MAny-Task computing

execution fabRIc at eXascales

Ke WangData-Intensive Distributed Systems Laboratory

Computer Science DepartmentIllinois Institute of Technology

February 14th, 2012

Page 2: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Acknowledgements

• DataSys Laboratory• Dr. Ioan Raicu• Juan Carlos Hernández

Munuera, MS 2011 • Hui Jin, Tonglin Li• Paper submission:

– Ke Wang, Ioan Raicu. “SimMatrix: Exploring Many-Task Computing through Simulations at Exascales”, under review at ACM HPDC 2012SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 2

Page 3: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 3

Page 4: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 4

Page 5: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Distributed Systems

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 5

Page 6: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

0

50

100

150

200

250

300

2004 2006 2008 2010 2012 2014 2016 2018

Nu

mb

er

of

Co

res

0

10

20

30

40

50

60

70

80

90

100

Ma

nu

fac

turi

ng

Pro

ce

ss

Number of CoresProcessing

Pat Helland, Microsoft, The Irresistible Forces Meet the Movable Objects, November 9th, 2007

Manycore Computing

• Today (2011): Multicore Computing– O(10) cores commodity architectures– O(100) cores proprietary architectures– O(1000) GPU hardware threads

• Near future (~2018): Manycore Computing– ~1000 cores/threads commodity architectures

6SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales

Page 7: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Exascale Computing

Top500 Performance Development,

http://top500.org/static/lists/2011/11/TOP500_201111_Poster.pdf 7

• Today (2012): 10 Petaflop Computing– O(100K) nodes (100X in the last 10 years) – O(1M) cores (1000X in the last 10 years)

• Near future (~2018): Exaflop Computing– ~1M nodes (10X) – ~1B processor-cores/threads (1000X)

Page 8: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Major Challenges of Exascale Computing

• Concurrency– Parallel programmability

• Resilience– MTTF decreases, MPI suffers

• I/O and Memory– Minimizing data movement

• Heterogeneity– Accelerators, GPUs, MIC

• Energy– 20MW limitation

8SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales

Page 9: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

MTC: Many-Task Computing

Number of Tasks

Input Data Size

Hi

Med

Low1 1K 1M

HPC(Heroic

MPI Tasks)

HTC/MTC(Many Loosely Coupled Tasks)

MapReduce/MTC(Data Analysis,

Mining)

MTC(Big Data and Many Tasks)

• Bridge the gap between HPC and HTC

• Applied in clusters, grids, and supercomputers

• Loosely coupled apps with HPC orientations

• Many activities coupled by file system ops

• Many resources over short time periodsSimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 9

Page 10: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

MTC Middleware

• Falkon– Fast and

Lightweight Task Execution Framework

– http://datasys.cs.iit.edu/projects/Falkon/index.html

• Swift– Parallel

Programming System

– http://www.ci.uchicago.edu/swift/index.php

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 10

Page 11: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 11

Page 12: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Long-Term Aims

• Address major exascale computing challenges:– Concurrency– Resilience– I/O and Memory– Heterogeneity

• Explore techniques to enable MTC at exascales• Design, Analyze, and Implement a distributed data-aware

execution fabric (MATRIX) supporting HPC/MTC workloads• Integrate MATRIX with parallel programming systems (e.g.

Swift, Charm++, MapReduce) and with the FusionFS distributed file system

• Prove that MTC applications can scale to exascales

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 12

Page 13: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

This Work’s Contributions

• Explore techniques to enable MTC to scale to exascales– Design, Analyze, and Implement a discrete-event

simulator (SimMatrix) enabling the study of MATRIX at extremely large scales (e.g. exascales)

– Identified work stealing as a viable technique to achieve load balance at exascales

– Provide evidence that work stealing is scalable by identifying optimal parameters affecting the performance of work stealing

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 13

Page 14: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 14

Page 15: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales

OverviewJob Scheduling Systems

• Efficiently manage the distributed computing power of workstations, servers, and supercomputers in order to maximize job throughput and system utilization.– Load balancing is critical

• Different scheduling strategies– Centralized scheduling hinders the scalability– Hierarchical scheduling has long job turnaround time – Distributed scheduling is a promising approach at exascales

• Work Stealing – a distributed scheduling strategy – Starved processors steal tasks from overloaded ones– Various parameters affect performance:

• Number of tasks to steal• Number of neighbors• Static or Dynamic random neighbors

15

Page 16: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

SimMatrix Architecture

Client

Submit tasks

Submit tasks

ClientArbitrary Node

Figure 1: Simulation architectures; the left part is the centralized one with a single dispatcher connecting all nodes, the right part is the homogeneous distributed topology with each node having the same number of cores and neighbors

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 16

Dispatcher

Page 17: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Simulations

• Continuous time simulations– Abandoned the idea of creating a separate thread

per simulated node: we found that on our 48-core system with 256GB of memory, we were limited to 32K threads

• Discrete event simulations– The only viable approach (today) to explore

scheduling techniques at exascales (millions of nodes and billions of cores)

– Created a unique object per simulated node, and converted any behavior to an event

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 17

Page 18: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 18

Page 19: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

At the Heart of SimMatrixGlobal Event Queue

Figure 2: Event State Transition DiagramSimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 19

• All events are inserted to the queue, sorted based on the occurrence time ascending

• Handle the first event, advance the simulation time and update the event queue

• Implemented as red-black tree based “TreeSet” in Java, which ensures Θ(log ) 𝑛time for insert & remove

Page 20: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Simulator Features

• Node load information– Nested hash maps provides extremely fast

performance at large scales• Dynamic Task Submission

– Aims to reduce the memory foot-print• Dynamic Poll interval

– Exponential backoff to reduce the number of messages and increase speed of simulation

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 20

Page 21: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Implementation

• SimMatrix is developed in JAVA– Sun 64-bit JDK version 1.6.0_22– 1500 lines of code– Code accessible at:

• http://datasys.cs.iit.edu/projects/SimMatrix/index.html

• SimMatrix has no other dependencies

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 21

Page 22: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 22

Page 23: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Experiment Environment

• Fusion system:– fusion.cs.iit.edu– 48 AMD Opteron cores at 1.93GHz– 256GB RAM– 64-bit Linux kernel 2.6.31.5– Sun 64-bit JDK version 1.6.0_22

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 23

Page 24: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Metrics

• Throughput– Number of tasks finished per second. Calculated as

total-number-of-tasks/simulation-time. • Efficiency

– The ratio between the ideal simulation time of completing a given workload and the real simulation time. The ideal simulation time is calculated by taking the average task execution time multiplied by the number of tasks per core.

• Load Balancing– We adopted the coefficient variance of the number of tasks finished by each

node as a measure the load balancing. The smaller the coefficient variance, the better the load balancing is. It is calculated as the standard-deviation/average in terms of number of tasks finished by each node.

• Scalability– Total number of tasks, number of nodes, and number of cores supported.

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 24

Page 25: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Workloads

• Synthetic workloads: – Uniform distributions with different average task

lengths, such as 10s (ave_10), 100s (ave_100), 1000s (ave_1000), 5000s (ave_5000), 10000s (ave_10000), and 100000s (ave_100000); also all tasks of 1 sec each (all_1)

• Realistic application workloads: – General MTC workload from 2008-2009 trace of

173M tasks; average task length 64±486s (mtc_64), using Gamma Distribution

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 25

Page 26: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales

Validation

Validate SimMatrix against the state-of-the-art MTC systems (e.g. Falkon), to ensure that the simulator can accurately predict the performance of current petascale systems. 26

Page 27: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales

Comparing Work Stealing to Falkon’s Naïve Distributed Scheduler

27

Fine grained workloads:• 2% 99.3%

efficiency increase

Coarse grained workloads:• 99%

99.999% efficiency increase

Page 28: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Scalability1M Nodes and 10B tasks

Memory consumption• <13 KB/task• <200 GB

CPU Time• <90 us/task• <260 hours

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 28

Page 29: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Scalability1M Nodes and 10B tasks

Efficiency• 90%+

Co-variance• <0.06• Load

imbalance of <600 tasks from 10K tasks per node

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 29

Page 30: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Work Stealing ParametersNumber of Tasks to Steal

30

Stealing half of neighbor’s work is best

strategy!0%

10%20%30%40%50%60%70%80%90%

100% No. of Tasks to Stealsteal_1steal_2steal_logsteal_sqrtsteal_half

No. of Nodes

Effici

ency

Page 31: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Work Stealing ParametersNumber of Neighbors (Static)

31

Requires linear number of neighbors for good

performance!

0%10%20%30%40%50%60%70%80%90%

100% No. of Static Neighbors

nb_2nb_lognb_sqrtnb_eighthnb_quarnb_half

No. of Nodes

Effici

ency

Page 32: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Work Stealing ParametersNumber of Neighbors (Dynamic Random)

32

An increasing number of neighbors are needed for 90%+ efficiency, with the largest scales requiring square root neighbors (e.g. 1K

neighbors from 1M nodes!0%10%20%30%40%50%60%70%80%90%

100% No. of Dynamic Random Neighbors

nb_1nb_2nb_lognb_sqrt

No. of Nodes

Effici

ency

Page 33: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Work Stealing ParametersOptimal Parameters Generality

33

The same optimal parameters achieve 90%+ efficiency across many different

workloads!0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100% Different Workloads

ave_1064+/-486ave_100ave_1000

No. of Nodes

Effici

ency

Page 34: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Work StealingThroughput

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 34

Centralized scheduling has severe bottleneck, especially for workload with fine granularity. Distributed scheduling has great scalability, for workload with coarse granularity, there is no obvious upper bound

0.125

0.5

2

8

32

128

512

2048

8191.99999999999

32767.9999999999

131072

524287.999999998

2097151.99999999

8388607.99999998

33554431.9999998

134217727.999999

Centralized(ave_5000)Distributed(ave_5000)Centralized(all_1)Distributed(all_1)

No. of Nodes

Thro

ughp

ut(t

asks

/sec

)

0

100000

200000

300000

400000

500000

600000

700000

Ave No. of Messages / tasks

No. of Nodes

Ave

rage

No

. of

Mes

sage

s pe

r Ta

sk

Page 35: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Load Balancing Visualization1024 Nodes and Ave_5000 Workload

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 35

Good Load Balancing

Square Root Dynamic Neighbors

Starvation

Square Root Static Neighbors

Good Load Balancing

Quarter Static Neighbors

Starvation

2 Static Neighbors

Page 36: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Summary Plot for Distributed Scheduling

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 36

Steady state utilization is ~100% at exascales

Page 37: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 37

Page 38: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Related Work

• Real Job Scheduling Systems: – Condor (University of Wisconsin), Bradley et al, 2012 – PBS (NASA Ames) , Corbatto et al, 2012 – LSF Batch (Platform Computing of Toronto), 2011– Falkon (University of Chicago), Raicu et al, SC07

• Job Scheduling System Simulators:– simJava (University of Edinburgh), Wheeler et al, 2004 – GridSim (University of Melbourne, Australia), Buyya et al, 2010

• Load Balancing: – Neighborhood averaging scheme, Sinha et al, 1993 – Charm++ (UIUC), Zheng et al, 2011

• Scalable Work Stealing– Dinan et al, SC09– Blumofe et al, Scheduling multithreaded computations by work stealing, 1994

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 38

Page 39: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 39

Page 40: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Contributions

• Designed, Analyzed, and Implemented a discrete-event simulator (SimMatrix) enabling the study of MTC workloads at exascales

• Identified work stealing as a viable technique to achieve load balance at exascales

• Provided evidence that work stealing is scalable by finding optimal parameters affecting the performance of work stealing– Number of tasks to steal is half– Dynamic random neighbors strategy is required– There must be a squared root number of neighbors

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 40

Page 41: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Outline

• Introduction & Motivation• Long-Term Aims and Contributions• SimMatrix Architecture• Implementation• Evaluation• Related Work• Contributions• Future Work & Conclusion

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 41

Page 42: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Future Work

• Explore work stealing for manycore processors with 1000 cores

• Enhancing the network topology model to allow complex networks

• Insight from SimMatrix will be used to develop MATRIX, a distributed task execution fabric– MATRIX will employ work stealing for distributed load

balancing– MATRIX will be integrated with other projects, such as

Swift (a data-flow parallel programming systems) and FusionFS(a distributed file systems)

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 42

Page 43: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

Conclusion

• Exascale systems bring great opportunities in unraveling of significant scientific mysteries

• There are significant challenges to achieve exascales, such as concurrency, resilience, I/O and memory, heterogeneity, and energy

• MTC requires a highly scalable and distributed task/job management system at large scales– Distributed scheduling is likely an efficient way to achieve

load balancing, leading to high job throughput and system utilization

• Work stealing is a scalable method to achieve load balance at exascales given the optimal parameters

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 43

Page 44: DataSys Laboratory Dr. Ioan Raicu Juan Carlos Hernández Munuera, MS 2011 Hui Jin, Tonglin Li Paper submission: –Ke Wang, Ioan Raicu. “SimMatrix: Exploring.

• More information:– http://datasys.cs.iit.edu/~kewang/ – http://datasys.cs.iit.edu/projects/SimMatrix/

• Contact:– [email protected]

• Questions?

More Information

SimMatrix: SIMulator for MAny-Task computing execution fabRIc at eXascales 44