Introduc)on to Cloud Compu)ng Dr. Zhenlin Wang Michigan Tech Very - - PowerPoint PPT Presentation

▶
introduc on to cloud compu ng
SMART_READER_LITE
LIVE PREVIEW

Introduc)on to Cloud Compu)ng Dr. Zhenlin Wang Michigan Tech Very - - PowerPoint PPT Presentation

Mar 20, 2023 128 likes •436 views

Introduc)on to Cloud Compu)ng Dr. Zhenlin Wang Michigan Tech Very Short Bio BS, MS, Peking University How did I get there? Why CS? PhD, University of MassachuseAs, Amherst Professor at Tech 2 Hobbies? Teaching and

slide-1
SLIDE 1

Introduc)on to Cloud Compu)ng

  • Dr. Zhenlin Wang

Michigan Tech

slide-2
SLIDE 2

Very Short Bio

  • BS, MS, Peking University

– How did I get there? – Why CS?

  • PhD, University of MassachuseAs, Amherst
  • Professor at Tech

2

slide-3
SLIDE 3

Hobbies?

  • Teaching and Research
  • Go
  • Tea & Sports Games

– Well, PE was the only course I couldn’t get an A

3

slide-4
SLIDE 4

Clouding CompuOng is here!

  • Google docs
  • Dropbox, Overleaf

– I am using them now

  • Tencent, TwiAer, Facebook

– Wechat: 600M users and counOng…

  • NeYlix, Amazon Prime

– I am a subscriber

  • ….

4

slide-5
SLIDE 5

What is Cloud Compu)ng?

5

Let’s hear from the “experts”

slide-6
SLIDE 6

What is Cloud Compu)ng?

6

The infinite wisdom of the crowds (via Google Suggest)

A few years back….

slide-7
SLIDE 7

What is Cloud CompuOng?

  • Now

7

slide-8
SLIDE 8

What is Cloud Compu)ng?

8

Larry Ellison, Co-founder, CEO of Oracle We’ve redefined Cloud CompuOng to include everything that we already do. . . . I don’t understand what we would do differently in the light of Cloud CompuOng other than change the wording of some of our ads.

slide-9
SLIDE 9

What is Cloud Compu)ng?

9

Richard Stallman GNU

It’s stupidity. It’s worse than stupidity: it’s a markeOng hype campaign

slide-10
SLIDE 10

What is Cloud Compu)ng?

10

Ron Rivest The R of RSA

Cloud CompuOng will become a focal point of our work in security. I’m

  • pOmisOc …
slide-11
SLIDE 11

What is Cloud Compu)ng?

11

It’s about jobs! It’s about small business!

slide-12
SLIDE 12

So, What really is Cloud Compu)ng?

Cloud compu)ng is a new compuOng paradigm, involving data and/or computaOon outsourcing, with

– Infinite and elasOc resource scalability – On demand “just-in-Ome” provisioning – No upfront cost … pay-as-you-go

12

That is, use as much or as less you need, use only when you want, and pay only what you use,

slide-13
SLIDE 13

NeYlix Version 1

13

NeDlix Home

Amazon.com Movies: Master copies

slide-14
SLIDE 14

14

What’s new in Today’s Clouds?

Besides massive scale, three major features: I. On-demand access: Pay-as-you-go, no upfront commitment.

– Anyone can access it (e.g., Washington Post – Hillary Clinton example)

II. Data-intensive Nature: What was MBs has now become TBs, PBs.

– Daily logs, forensics, Web data, photos, videos, etc. – Do you know the size of Wikipedia dump?

III. New Cloud Programming Paradigms: MapReduce/Hadoop, Pig LaOn, and many others.

– High in accessibility and ease of programmability

CombinaOon of one or more of these gives rise to novel and unsolved distributed compuOng problems in cloud compuOng.

slide-15
SLIDE 15

The real story

“CompuOng UOlity” – holy grail of computer science in the 1960s. Code name: MULTICS (MulOplexed InformaOon and CompuOng Service)

15

Why it failed?

  • Ahead of Ome … lack of communicaOon tech.

(In other words, there was NO (public) Internet)

  • And personal computer became cheaper and

stronger

slide-16
SLIDE 16

The real story

Mid to late ’90s, Grid compu)ng was proposed to link and share compuOng resources

16

slide-17
SLIDE 17

The real story … conOnued

17

Post-dot-com bust, big companies ended up with large data centers, with low uOlizaOon Solu)on: Throw in virtualizaOon technology, and sell the excess compuOng power And thus, Cloud Compu)ng was born …

slide-18
SLIDE 18

Cloud compuOng provides numerous economic advantages

For clients:

– No upfront commitment in buying/leasing hardware – Can scale usage according to demand – Barriers to entry lowered for startups

For providers:

– Increased uOlizaOon of datacenter resources

18

slide-19
SLIDE 19

Cloud compuOng means selling “X as a service”

IaaS: Infrastructure as a Service

– Selling virtualized hardware

PaaS: PlaYorm as a service

– Access to a configurable plaYorm/API

SaaS: Somware as a service

– Somware that runs on top of a cloud

19

slide-20
SLIDE 20

Cloud compuOng architecture

20

e.g., Web browser SaaS , e.g., Google Docs PaaS, e.g., Google AppEngine IaaS, e.g., Amazon EC2

slide-21
SLIDE 21

Top 10 Obstacles (Berkley’09)

  • Availability of Service

– Use MulOple Cloud Providers; Use ElasOcity to Prevent DDOS

  • Data Lock-In

– Standardize APIs – CompaOble SW to enable Surge CompuOng

  • Data ConfidenOality and Auditability

– Deploy EncrypOon, VLANs, Firewalls; Geographical Data Storage

  • Data Transfer BoAlenecks

– FedExing Disks; Data Backup/Archival; Higher BW Switches

  • Performance Unpredictability

– I/O interferences – Improved VM Support; Flash Memory; Gang Schedule VMs

21

slide-22
SLIDE 22

Top 10 Obstacles

  • Scalable Storage

– Invent Scalable Store

  • Bugs in Large Distributed Systems

– Invent Debugger that relies on Distributed VMs

  • Scaling Quickly

– Invent Auto-Scaler that relies on machine learning – Snapshots for ConservaOon

  • ReputaOon Fate

– Sharing offer reputaOon-guarding services like those for email

  • Somware Licensing

– Pay-for-use licenses; Bulk use sales

22

slide-23
SLIDE 23

My Research

  • Memory system modeling and virtualizaOon
  • Dynamic data center resource management

23

slide-24
SLIDE 24

Memory Balancing

  • Dynamic member

balancing for virtual machines (Zhao&Wang VEE’99, Wang et al. ATC’11)

24

2G? 2G? 2G?

slide-25
SLIDE 25

Memory Balancing

25

2G? 2G? 2G?

473.astar

slide-26
SLIDE 26

Memory Balancing: Demand PredicOon

26

L1,L2,DTLB Monitoring AVL-tree Based LRU Dynamic Hot Set restore revoke Phase detectioin Miss ratio curve WSS Estimation Intermittent Memory Tracking resize resize Hardware Kernel Control Plane

slide-27
SLIDE 27

Key-Value Store Management

  • LAMA: Op(mized Locality-aware Memory

Alloca(on for Key-value Cache (Hu et al. ATC 15)

27

How to dynamically adjust cache allocaOon? class A class B

slide-28
SLIDE 28

Cross-Architecture Co-Tenancy PredicOon

  • NSF CSR’14 with Dr. Laura Brown (CCGRID’15,

AAAI PhD ConsorOum’15)

28

Batch’s Pressure Score Core 1 Core 2 Shared cache/Memory Core 1 Core 2 Shared cache/Memory Hardware Configuration1(HW1) ... ... ... y=f_astar(x) y=f_gcc(x) ... ... ... y=g_astar(x) y=g_gcc(x) ... ... ... y=f_astar(x) y=f_gcc(x) ... ... ... y=g_astar(x) y=g_gcc(x) q_astar ... ... ... q_gcc p_astar ... ... ... p_gcc LatencySensitive Programs Batch programs as interference Profiling Hardware Configuration2(HW2) Sensitivity Curve Programs Batch programs as interference LatencySensitive report pressure score of batch program Profiling Sensitivity Curve ... ... astar gcc astar gcc Curve fitting Curve fitting Training Regression cross architectural mapping g_program=func(f_program,HW1,HW2) Model: p_program=func’(q_program,HW1,HW2) Model: Training Regression Input Input latencysensitive program A’s sensitivity function on HW1 batch program B’s pressure score

  • n HW1

Output Output Program A’s sensitivity curve

  • n HW2

Program B’s pressure score

  • n HW2

pressure score performace degradation final prediction

Performance?

slide-29
SLIDE 29

Systems research is exciOng! Students are always welcome!

– Junior year is the best Ome to join

29