pricing for utility-driven resource management and allocation in clusters chee shin yeo and rajkumar...
TRANSCRIPT
![Page 1: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/1.jpg)
Pricing for Utility-driven Resource Management
and Allocation in Clusters
Chee Shin Yeo and Rajkumar Buyya
Grid Computing and Distributed Systems (GRIDS) Lab. Dept. of Computer Science and Software EngineeringThe University of Melbourne, Australia
www.gridbus.org/
WW Grid
![Page 2: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/2.jpg)
2
Presentation Outline
Motivation Computation Economy Economy-based Admission Control,
Resource Allocation & Job Control Pricing Function Performance Evaluation Conclusion and Future Work
![Page 3: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/3.jpg)
3
Motivation
Cluster-based systems have gained popularity and widely adopted 75% of Top500 supercomputers world-
wide based on Cluster architecture. Clusters are used in not only used in
scientific computing, but also in driving many commercial applications.
Many Corporate Data Centers are cluster-based systems.
![Page 4: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/4.jpg)
4
Problem and our Proposal
However, RMS responsible for managing clusters and allocating resources to users
Still adopts system-centric approaches such as FCFS with some static pariorities.
Maximize CPU throughput & CPU utilization Minimize average waiting time & average response time
They provide no or minimal means for users to define Quality-Of-Service (QoS) requirements.
We propose the use of user-centric approaches such as computational economy in management of cluster resources.
![Page 5: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/5.jpg)
5
Computational Economy
Management of shared resources with economic accountability is effective: Regulates supply and demand of cluster
resources at market equilibrium User-centric management of clusters
Users express Quality Of Service (QoS) requirements
Users express their valuation for the required service
Economic incentives for both users and cluster owner as a means of feedback
![Page 6: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/6.jpg)
6
Utility-driven Cluster RMS Architecture
![Page 7: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/7.jpg)
7
Economy-based Admission Control & Resource Allocation
Uses the pricing function to compute cost for satisfying the QoS of a job as a means for admission control
Regulate submission of workload into the cluster to prevent overloading
Provide incentives Deadline -- $ Execution Time -- $ Cluster Workload -- $
Cost acts as a mean of feedback for user to respond to
![Page 8: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/8.jpg)
8
Economy-based Admission Control & Resource Allocation
Accept or reject based on 3 criteria (consider required QoS) resource requirements that are needed by
the job to be executed deadline that the job has to be finished budget to be paid by the user for the job
to be finished within the deadline Requires estimated execution time Allocates job to node with least
remaining free processor time
![Page 9: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/9.jpg)
9
Job Control: Economy-based Proportional Resource Sharing
Monitor and enforce required deadline.
Time-shared Allocate resources proportional to the
needs of jobs based on the estimated execution time and required deadline
Update processor time partition periodically
![Page 10: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/10.jpg)
10
Essential Requirements for Pricing
Flexible Easy configuration
Fair Based on actual usage
Dynamic Not static
Adaptive Changing supply and demand of resources
![Page 11: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/11.jpg)
11
Pricing Function
![Page 12: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/12.jpg)
12
Pricing Function
![Page 13: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/13.jpg)
13
Processing Cost Functions for Different Scheduling Algorithms
First-Come-First-Served (FCFS)
Economy based Proportional Resource Sharing (Libra)
Libra with dynamic pricing (Libra+$)
![Page 14: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/14.jpg)
14
Performance Evaluation: Simulation
Simulation Model Simulated scheduling for a cluster
computing environment using the GridSim toolkit (http://www.gridbus.org/gridsim)
Simulated Cluster manjra.cs.mu.oz.au (13 single-processor
nodes with Pentium4 2-GHz CPU)
![Page 15: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/15.jpg)
15
Experimental Methodology
![Page 16: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/16.jpg)
16
Evaluation Metrics
Job QoS Satisfaction Cluster Profitability Average Waiting Time Average Response Time
![Page 17: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/17.jpg)
17
Normalised Comparison of FCFS, Libra & Libra+$
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Job QoSSatisfaction
Cluster Profitability Average WaitingTime
Average ResponseTime
FCFS
Libra
Libra+$, β = 0.01
![Page 18: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/18.jpg)
18
Varying Cluster Workload
Scheduling policies First-Come-First-Served (FCFS) Economy based Proportional Resource
Sharing (Libra) Libra with dynamic pricing (Libra+$)
An increasing mean job execution time 6, 7, 8, 10, 15 and 30 hours
![Page 19: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/19.jpg)
19
Impact of Increasing Job Execution Time on Job QoS Satisfaction
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
6 7 8 10 15 30
Mean Job Execution Time (hours)
Job
Qo
S S
ati
sfa
ctio
n (
%)
FCFS
Libra
Libra+$, β = 0.01
![Page 20: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/20.jpg)
20
Impact of Increasing Job Execution Time on Cluster Profitability
0%
5%
10%
15%
20%
25%
30%
6 7 8 10 15 30
Mean Job Execution Time (hours)
Clu
ster
Pro
fita
bili
ty (
%)
FCFS
Libra
Libra+$, β = 0.01
![Page 21: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/21.jpg)
21
Varying Pricing Factor for Different Level of Sharing
Scheduling policies Libra with dynamic pricing (Libra+$)
An increasing dynamic pricing factor β 0.01, 0.1, 0.3, and 1
![Page 22: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/22.jpg)
22
Impact of Increasing Dynamic Pricing Factor on Job QoS Satisfaction
0%
10%
20%
30%
40%
50%
60%
70%
80%
0.01 0.1 0.3 1
Dynamic Pricing Factor β
Job
Qo
S S
ati
sfa
ctio
n (
%)
FCFS
Libra
Libra+$
![Page 23: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/23.jpg)
23
Impact of Increasing Dynamic Pricing Factor on Cluster Profitability
0%
10%
20%
30%
40%
50%
60%
70%
80%
0.01 0.1 0.3 1
Dynamic Pricing Factor β
Clu
ster
Pro
fita
bili
ty (
%)
FCFS
Libra
Libra+$
![Page 24: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/24.jpg)
24
Tolerance against Estimation Error
Under-estimated execution time EEi e.g. job whose execution time Ei = 60 hours
has EEi = 30 hours for estimation error = 50% Scheduling policies
Libra – Economy based Proportional Resource Sharing (Libra)
Libra with dynamic pricing (Libra+$) An increasing estimation error for
estimated execution time EEi 0%, 10%, 30% and 50%
![Page 25: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/25.jpg)
25
Impact of Increasing Estimation Error on Job QoS Satisfaction
0%
10%
20%
30%
40%
50%
60%
70%
80%
0% 10% 30% 50%
Estimation Error for Estimated Execution Time EE i (%)
Job
Qo
S S
ati
sfa
ctio
n (
%)
Libra
Libra+$, β = 0.01
Libra+$, β = 0.1
Libra+$, β = 0.3
Libra+$, β = 1.0
![Page 26: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/26.jpg)
26
Impact of Increasing Estimation Error on Cluster Profitability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0% 10% 30% 50%
Estimation Error for Estimated Execution Time EE i (%)
Clu
ste
r P
rofi
tab
ility
(%
)
Libra
Libra+$, β = 0.01
Libra+$, β = 0.1
Libra+$, β = 0.3
Libra+$, β = 1.0
![Page 27: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/27.jpg)
27
Conclusion & Future Work
Importance of effective pricing function (demand exceeds supply of resources)
Satisfy four essential requirements for pricing
Serves as means of admission control Tolerance against estimation errors Higher benefits for cluster owner Future work
Explore different pricing strategies Examine different application models
![Page 28: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/28.jpg)
Backup
![Page 29: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/29.jpg)
29
Related Work
Traditional cluster RMS Load Sharing Facility (LSF) – Platform Load Leveler – IBM Condor – University of Wisconsin Portable Batch System (PBS) – Altair Grid
Technologies Sun Grid Engine (SGE) – Sun Microsystems
Market-based cluster RMS REXEC Libra
![Page 30: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/30.jpg)
30
Job details eg. Estimated execution time
Resource requirements eg. Memory size, Disk storage size
QoS constraints eg. Deadline, Budget
QoS optimization eg. Time, Cost
User-level Job Submission Specification
![Page 31: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/31.jpg)
31
Job QoS Satisfaction
Performance Evaluation Metrics
![Page 32: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/32.jpg)
32
Cluster Profitability
Performance Evaluation Metrics
![Page 33: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/33.jpg)
33
Performance Evaluation Metrics
Average Waiting Time
![Page 34: Pricing for Utility-driven Resource Management and Allocation in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems (GRIDS)](https://reader036.vdocuments.mx/reader036/viewer/2022082805/5514647c550346284e8b5a91/html5/thumbnails/34.jpg)
34
Performance Evaluation Metrics
Average Response Time