Performance Modelling and Simulation of Service Chains for Telecom Clouds

(1)

Performance Modelling

and Simulation of Service Chains for Telecom Clouds

Michel Gokan Khan

Mic hel Gokan Khan | P erformance M odelling and Simulation of Service Chains for T elecom Clouds | 2021:14

Performance Modelling and Simulation of Service Chains for Telecom Clouds

New services and ever increasing traffic volumes require the next generation of mobile networks, e.g. 5G, to be much more flexible and scalable. The primary enabler for its flexibility is transforming network functions from proprietary hardware to software using modern virtualization technologies, paving the way of virtual network functions (VNF). Such VNFs can then be flexibly deployed on cloud data centers while traffic is routed along a chain of VNFs through software-defined networks. However, such flexibility comes with a new challenge of allocating efficient computational resources to each VNF and optimally placing them on a cluster.

In this thesis, we argue that, to achieve an autonomous and efficient performance optimization method, a solid understanding of the underlying system, service chains, and upcoming traffic is required. We, therefore, conducted a series of focused studies to address the scalability and performance issues in three stages.

We first introduce an automated profiling and benchmarking framework, named NFV-Inspector to measure and collect system KPIs as well as extract various insights from the system. Then, we propose systematic methods and algorithms for performance modelling and resource recommendation of cloud native network functions and evaluate them on a real 5G testbed. Finally, we design and implement a bottom-up performance simulator named PerfSim to approximate the performance of service chains based on the nodes’ performance models and user-defined scenarios.

LICENTIATE THESIS | Karlstad University Studies | 2021:14 Faculty of Health, Science and Technology

Computer Science LICENTIATE THESIS | Karlstad University Studies | 2021:14

ISSN 1403-8099

ISBN 978-91-7867-209-7 (pdf)

ISBN 978-91-7867-199-1 (print)

(2)

LICENTIATE THESIS | Karlstad University Studies | 2021:14

Performance Modelling

and Simulation of Service Chains for Telecom Clouds

Michel Gokan Khan

(3)

Print: Universitetstryckeriet, Karlstad 2021 Distribution:

Karlstad University

Faculty of Health, Science and Technology

Department of Mathematics and Computer Science SE-651 88 Karlstad, Sweden

+46 54 700 10 00

©

The author ISSN 1403-8099

urn:nbn:se:kau:diva-83687

Karlstad University Studies | 2021:14 LICENTIATE THESIS

Michel Gokan Khan

Performance Modelling and Simulation of Service Chains for Telecom Clouds

WWW.KAU.SE

ISBN 978-91-7867-209-7 (pdf)

ISBN 978-91-7867-199-1 (print)

(4)

P S S

Michel Gokan Khan

Department of Mathematics and Computer Science, Karlstad University, Sweden

G

K

G

M

(5)

(6)

K

M

K K G

K M K K

K M G K

(7)

(8)

P

K

IEEE 7th International Conference on Cloud Network-

ing (CloudNet 2018) I S S

K M

IEEE 4th Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN 2018)

I P

M K

K M

M IEEE 6th International Conference on

Network Softwarization (NetSoft 2020)

I S P

K K

M IEEE 11th International Confer-

ence on Networks of the Future (IEEE NoF 2020)

K Under review

S

P

P I

P II

M

(9)

P III

K P I

P

M K K

K

Mobile Networks and Applications

K Under submission

P M M

7th IEEE International Conference on E-Health and Bioengineering (EHB 2019)

(10)

M International Conference onG Interfaces and Human Computer Interaction 2019

M M

Interactivity and the Future of the Human-Computer Interface K G G

M K

Edge Computing: Models, technologies and applications

M K

Edge Computing: Models, technologies and applications

Edge Computing:

Models, technologies and applications

(11)

(12)

P

Introductory Summary

I

S P

Paper I

I S P

I

I S

I

(13)

K

Paper II

I P

I I I

Paper III

P S

I

P S

S MM

M M

(14)

Paper I

P P S

I S

M

P S

M K

K

S

Paper

P S P S

I

P S I

M

(15)

P S

S

MM

(16)

I S

Martin Reiser (1987)

(17)

(18)

I

Network Function Virtualization

Network Functions Virtual Machines M

Containers Virtual Network Function

M 5G core

G G microservice architecture

G cloud na-

tive modern, dynamic environ-

ments such as public, private, and hybrid clouds decoupling of application con- tainers from the details of the systems on which they run

service chains

Cloud Native Network Function

S

(19)

P S

Network Function Virtualization Infras- tructure

S

G

M

(20)

S

G

(21)

K M K

M M

K

M M

K M

(22)

P

Key Performance Indicators K

K

Statistical Machine Learning M

Invasive Non-invasive

M Server-

less Application Analytics Framework Function-as-a- Service

(23)

M

P

Captial Expendi-

tures Operational Expenses Service Providers

Service Level Agree- ment

M rich statistical performance models

policy simulator

model management M

(24)

M

S

(25)

(26)

How can we systematically analyze and profile large-scale systems?

K K

How can we accurately model and simulate the performance of large-scale service chains?

How can we optimally allocate resources to components of service chains and place them in a cluster while guaranteeing QoS constraints?

(27)

A framework for systematic profiling of services and analysis of involved KPIs M

P I

P II

P I

K

Two distinct approaches for systematic performance modelling

P III

P K

(28)

An SLA-aware initial resource recommendation method for cloud native sys- tems

P III

M

G

A performance simulator for cloud native computing

P

M

∼

(29)

computer science

mathematical scientific engineering design process

Deploying on PTE Performance Testing and Monitoring

Performane

Analysis Performance Modeling Modeling Procedure

1 2

4

3 All models

extracted Performance

simulation 5

Dynamic Control

Deploying in PE

Dynamic Monitoring

7 8

System and Traffic Analysis

9

Dynamic Policy Tunning

Current scope Future work

Finding an optimal resource allocation and placement on PE

6

K

(30)

1 2

4 5 3 Stages

Objectives Papers

A B C D E

I II III IV 6 V

(31)

S P

P I I S P

G

P II P

I

K K

P III P S

G G

P I P

P S

K

(32)

K

P P S P S

K

∼

G G G

G

(33)

M IEEE Communications Magazine

M IEEE/ACM

Trans. Netw.

2014 IEEE International Advance Computing Con- ference (IACC)

M M

International Conference on Optical Network Design and Modeling (ONDM)2020

2016 13th IEEE Annual Con- sumer Communications Networking Conference (CCNC)

G M

Proceedings of the 2009 Conference on Hot Topics in Cloud Computing Foundations of software and system performance engineering:

process, performance modeling, requirements, testing, scalability, and practice

M K

NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Sympo- sium

K K

2015 IEEE First International Conference on Big Data Computing Service and Applications

2017 IEEE Conference on ComputerM Communications Workshops (INFOCOM WKSHPS)

(34)

M Future Generation Computer Systems

G M

2017 Winter Simulation Conference (WSC) M

Software: Practice and Experience

Software: Practice and experience

2015 IEEE Confer- ence on Network Function Virtualization and Software Defined Network (NFV- SDN)

2012 IEEE 8th International Confer- ence on E-Science

https://github.com/cncf/toc/

blob/master/DEFINITION.md

Engineering design methods : strategies for product design

M 2015 IEEE Conference on Network Function Virtualization and Software Defined Network (NFV-SDN)

M Philosophy of science : the central issues

G

IEEE Transactions on Services Computing

2013 39th Euromicro Conference on Software Engineering and Advanced Applications

(35)

2013 IEEE/ACM 6th International Conference on Utility and Cloud Comput- ing

M 2015

IEEE 23rd International Conference on Network Protocols (ICNP) G IEEE Transactions on Network and Service Management

G K G

Software: Practice and Experience

M M

ICC 2019 - 2019 IEEE International Conference on Com- munications (ICC)

M M M M

Simulation Modelling Practice and Theory

G K G M K

Journal of Systems and Software

K K

2015 17th Asia-Pacific Network Operations and Management Sym- posium (APNOMS)

K K G

2010 IEEE Global Telecommunications Conference GLOBECOM 2010

IEEE Communications Surveys Tutorials

G M K

M G G G

2015 11th International Conference on Network and Service Management (CNSM)

K Computer

Networks

(36)

G K K M

2009 IEEE International Conference on Cluster Computing and Workshops

M G M M

IEEE Access

G M Chinese Journal of Electronics

K M

2016 IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN)

Proceedings of the Symposium on SDN

Research M

G K

2008 International Symposium on a World of Wireless, Mobile and Multimedia Networks

2020 IEEE Wireless Commu- nications and Networking Conference (WCNC)

M G G

Journal of Grid Computing M

M K

Software-Defined Networks (EWSDN), 2016 Fifth European Workshop on

G How to solve it : a new aspect of mathematical method

M G M M M M M

Simulation Modelling Practice and TheoryM M

G The ns-3 Network Simulator

(37)

IEEE Com- munications Magazine

G IEEE INFOCOM 2019 - IEEE

Conference on Computer Communications

M K K G

Electronics

K K G

G Machine Learning, Image Processing, Network Security and Data Sciences

M M M M M

2017 IFIP/IEEE Symposium on Integrated Network and Service Management (IM)

2015 15th IEEE/ACM International Symposium on Cluster, Cloud andM Grid Computing

Transactions on Emerging Telecommunications Technologies

M K

Methods and Applications for Modeling and Simulation of Complex Sys- tems

M M M

SAE Technical Paper

2018 IEEE Intl Conf on Parallel Distributed Processing with Applications, Ubiquitous Com- puting Communications, Big Data Cloud Computing, Social Computing Net- working, Sustainable Computing Communications (ISPA/IUCC/BDCloud/So- cialCom/SustainCom)

(38)

M G M

2016 45th Interna- tional Conference on Parallel Processing Workshops (ICPPW)

K G G

Journal of Systems Architecture

(39)

(40)

Performance Modelling

and Simulation of Service Chains for Telecom Clouds

Michel Gokan Khan

Mic hel Gokan Khan | P erformance M odelling and Simulation of Service Chains for T elecom Clouds | 2021:14

Performance Modelling and Simulation of Service Chains for Telecom Clouds

New services and ever increasing traffic volumes require the next generation of mobile networks, e.g. 5G, to be much more flexible and scalable. The primary enabler for its flexibility is transforming network functions from proprietary hardware to software using modern virtualization technologies, paving the way of virtual network functions (VNF). Such VNFs can then be flexibly deployed on cloud data centers while traffic is routed along a chain of VNFs through software-defined networks. However, such flexibility comes with a new challenge of allocating efficient computational resources to each VNF and optimally placing them on a cluster.

In this thesis, we argue that, to achieve an autonomous and efficient performance optimization method, a solid understanding of the underlying system, service chains, and upcoming traffic is required. We, therefore, conducted a series of focused studies to address the scalability and performance issues in three stages.

We first introduce an automated profiling and benchmarking framework, named NFV-Inspector to measure and collect system KPIs as well as extract various insights from the system. Then, we propose systematic methods and algorithms for performance modelling and resource recommendation of cloud native network functions and evaluate them on a real 5G testbed. Finally, we design and implement a bottom-up performance simulator named PerfSim to approximate the performance of service chains based on the nodes’ performance models and user-defined scenarios.