IaaS Cloud Architectures: Virtualized Data Centers to Federated Cloud Infrastructures
iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers
description
Transcript of iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers
![Page 1: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/1.jpg)
iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers
Hui Zhang1 , Kenji Yoshihira1, Ya-Yunn Su2 , Guofei Jiang1, Ming Chen3, Xiaorui Wang3
1. NEC Laboratories America2. National Taiwan University3. University of Tennessee
![Page 2: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/2.jpg)
Virtualized data centers: server consolidation and green IT
Resource Pool
• Server consolidation - virtualization facilitates consolidation of several physical servers onto a single high end system
— Reduces management costs/overheads — Increases overall utilization
• Green IT - computing more, consume less— Improving infrastructure efficiency —Increasing IT productivity
Today Future
DCiE =IT load power
Total data centerInput power
DCpW =
Data center useful work
Total facility power
DCiE: Data center infrastructure efficiency DCpW: Data center performance per Watt
iPOEM 2ICAC2011
![Page 3: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/3.jpg)
Virtualized data center management
• Server utilization based performance and power management mechanisms– VMware DPM, NEC SSC, IBM Tivoli…
CP
U u
tiliz
atio
n
CPUlow
Overload threshold
CPUhigh
Power-savingmode
iPOEM 3ICAC2011
Management Configuration
![Page 4: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/4.jpg)
iPOEM: a middleware for integrated power and performance management
• Features declarative management methodology1. accepts higher-level management objectives
• target system status set on individual management components
2. generates low-level management configurations.• Configuration settings of individual management components
• A GPS tool is a good metaphor.
operatordriver
carGPS device
• How can I go to NYC?• How can I go to NYC?
1. Map2. Direction1. Map2. Direction
iPOEM
• How can I get the system to 20% less power cost?
• How can I get the system to 20% less power cost?
1. System status2. Management
decisions
1. System status2. Management
decisions
system
iPOEM 4ICAC2011
![Page 5: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/5.jpg)
Research Goal
Data center administrator
Performance
management
Performance
management
Power manageme
nt
Power manageme
nt
Thermal manageme
nt
Thermal manageme
nt
Application
management
Application
management
t t t
t t t
t t t
t t t
t t t
t t t
System Dynamics
Workload Dynamics
failure remove/add
migration
Management challenges
Human-friendly management interfaces
iPOEM
System Complexity
Virtualized data center
![Page 6: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/6.jpg)
iPOEM APIs
iPOEM ICAC2011 6
API 1 : get_position() API 2 : put_position()
Input: Management Configuration
(Time start, Time end)
Workload (reshaping-scheme)
VM-server map, resource inventory
Output: System Status
March 2010Performance cost: server overloading time in percentage.
Performance cost: server overloading time in percentage.
time
CPU utilization
Lo
ad
th
resh
old
Lo
ad
th
resh
old
Power cost: KWatts, total
power consumed
Power cost: KWatts, total
power consumed
Operation cost:VM migrations
Operation cost:VM migrations
System status is described in 3 metrics
Input: Target Performance & Power
(Time start, Time end)
Workload (reshaping-scheme)
VM-server map, resource inventory
Output: Management Configuration
![Page 7: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/7.jpg)
iPOEM architecture
iPOEM 7ICAC2011
![Page 8: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/8.jpg)
iPOEM management configuration engine
8
iPOEM 8ICAC2011
API 1 : get_position()
Management Configuration
CPU High CPU High, Low
![Page 9: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/9.jpg)
System status as a function of management configurations
9
iPOEM 9ICAC2011
![Page 10: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/10.jpg)
Formal description of system status functions
10
Assume a homogeneous system, and the workload remains the same for different configuration settings.
Theorem 1. Performance-cost(CPUhigh) is a non-decreasing function of CPUhigh.
Theorem 2. Power-cost(CPUlow) is a non-increasing function of CPUlow.
iPOEM 10ICAC2011
![Page 11: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/11.jpg)
iPOEM configuration searching algorithm
11
• A O(logR) searching algorithmowhere R =CPUmax-CPUmin, the allowable load range
iPOEM 11ICAC2011
Binary search
![Page 12: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/12.jpg)
iPOEM prototype implementation
12
iPOEM 12ICAC2011
![Page 13: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/13.jpg)
iPOEM System positioning services
• Position reporting • Destination searching
iPOEM 13ICAC2011
![Page 14: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/14.jpg)
Evaluation: data center workload traces
• Traces on 2525 servers from 10 IT systems
– Each is regarded as a VM in the simulations.
• Monitoring data: CPU utilization. • 1 week length, 15 minute monitoring
frequency – 672 time points
iPOEM 14ICAC2011
![Page 15: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/15.jpg)
Evaluation: methodology • Run the iPOEM prototype as an offline engine.
– It is driven by the data traces stored in the monitoring database, and emulates the integrated management in a virtualized data center hosting the 2, 525 servers as VMs.
• The system and management configuration settings– Performance manager and power manager
• Implementation of the simplified schemes in NEC SigmaSystemCenter middleware
– The default <CPUlow,CPUhigh> setting is <40%, 80%>.– The physical servers are homogeneous with the same CPU specs
• 3GHZ Quadra-core (the most common CPU model in the traces).
– Performance cost: number of performance violation in a time epoch• A server has a performance violation at a time point when its CPU utilization is larger than a threshold
(90% in the paper).
– Power cost: we assume power consumption per server is either 0 (power-off mode) or 200Watts (power-on mode), simplified on the power model profiled in the local testbed.
– Operation cost: the number of VM migrations that the performance and power managers need to execute for the server load configuration enforcement.
iPOEM 15ICAC2011
![Page 16: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/16.jpg)
iPOEM engine performance
iPOEM engine response time to service requests
iPOEM 16ICAC2011
![Page 17: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/17.jpg)
iPOEM auto-piloting service• Sensitivity based optimization [Markovic et al. 2004]
17
where Pth is the upper bound of the performance cost.
iPOEM 17ICAC2011
![Page 18: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/18.jpg)
iPOEM auto-piloting evaluation
18
Comparison of Auto-piloting and three static configuration schemes
Auto-piloting management configuration evolutioniPOEM 18ICAC2011
![Page 19: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/19.jpg)
Conclusions & Future Work
• iPOEM, an integrated power and performance management middleware in an virtualized infrastructure.– human-friendly interfaces for multi-objective
management.
• Future work– Meta-management integrating more objectives.
• explosive growth of the system state space
– mashup services for customized tenant management• new API designs
iPOEM 19ICAC2011
![Page 20: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/20.jpg)
Thank you.
• Questions?
iPOEM ICAC2011 20
![Page 21: iPOEM: A GPS Tool for Integrated Management in Virtualized Data Centers](https://reader033.fdocuments.in/reader033/viewer/2022051623/56815a87550346895dc7f868/html5/thumbnails/21.jpg)
Appendix
iPOEM ICAC2011 21