A Computation- and Network-Aware Energy Optimization model for - - PowerPoint PPT Presentation

CLOSER 2017, April 24-26, Porto 1

A Computation- and Network-Aware Energy Optimization model for Virtual Machine Allocation

• C. Canali, R. Lancellotti

Department of Engineering “Enzo Ferrari”, University of Modena and Reggio Emilia

• M. Shojafar

Italian National Council for Telecommunications (CNIT)

CLOSER 2017, April 24-26, Porto 2

Motivation

• Energy consumption in Cloud

– Typical problem of server consolidation, but... – Network-related energy is often neglected – VMs migration: additional energy consumption

• Challenges of future Cloud systems

– Network softwarization: SDN → SDDC – Dynamic VMs behavior → VMs migrations

CLOSER 2017, April 24-26, Porto 3

Reference Scenario

• C
• m

p l e x n e t w

• r

k t

• p
• l
• g

y

• I

n t e r a c t i

• n

b e t w e e n n e t w

• r

k a n d a l l

• c

a t i

• n

m g r .

CLOSER 2017, April 24-26, Porto 4

Model

• Multi-dimensional bin packing problem
• Use of dynamic programming:

– Time divided in time slots – Start from placement at previous time slot – Define migrations of VMs – Turn ON/OFF servers

• G
• a

l s

– Minimize energy consumption – No parameters to tune

CLOSER 2017, April 24-26, Porto 5

Objective function

• Eobj: 3 components (in most complete form)
• Energy for computation EC
• Energy for data transfer ED
• Energy for VMs migrations EM

CLOSER 2017, April 24-26, Porto 6

Objective function

• Energy for computation EC
• Minimum energy

consumption for servers turned ON

• Linear dependence from

the CPU utilization of servers

CLOSER 2017, April 24-26, Porto 7

Objective function

• Energy for data transfer ED
• Minimum energy for network interfaces in servers

turned ON

• Linear dependence on data exchanged among

servers

• Captures network topology through parameter

Edi1,di2

CLOSER 2017, April 24-26, Porto 8

Objective function

• Energy for VMs migrations EM
• Computational overhead for servers
• Data transfer of VM memory

CLOSER 2017, April 24-26, Porto 9

Constraints

• Resource requests by VMs on a server must not

exceed server capacity:

– CPU – Memory – Network: no data exchange within the server

• VMs allocation only on servers turned ON

CLOSER 2017, April 24-26, Porto 10

Constraints

• Each VM is placed one and only one server
• Consistency of

VMs migrations

• Boolean nature of the problem

CLOSER 2017, April 24-26, Porto 11

Considered Alternatives

• Our proposal

– Migration Aware (MA)

Eobj=EC+ED+EM

• Existing solutions:

– No Migration Aware (NMA)

Eobj=EC+ED

– No Network Aware (NNA)

Eobj=EC

CLOSER 2017, April 24-26, Porto 12

Simulation Setup

• Resource requests from real VMs

– Default: 80 VMs

• Power consumption from datasheets
• Two network behavior scenarios:

– Network 1: Random interaction – Network 2: Pareto law interaction

(20% of destination IPs receive 80% of traffic)

• AMPL problem formulation

– CPLEX 12 solver

CLOSER 2017, April 24-26, Porto 13

Comparison

Network 1 Network 2

CLOSER 2017, April 24-26, Porto 14

Impact of Migration

Network 1 Network 2

CLOSER 2017, April 24-26, Porto 15

Results stability

CLOSER 2017, April 24-26, Porto 16

Conclusions

• Challenges of VMs placement in cloud

– Network becomes more important (SDDC) – More dynamic VMs behavior (migrations)

• Limitation of existing models
• → Migration-Aware model for VMs placement

– No parameter tuning required

• Future work:

– More focus on SDDC, model improvement – Heuristics for scalability

CLOSER 2017, April 24-26, Porto 17

A Computation- and Network-Aware Energy Optimization model for Virtual Machine Allocation Contact: Riccardo Lancellotti

riccardo.lancellotti@unimore.it