Intro 2 Cloud Computing

Published on May 2016 | Categories: Documents | Downloads: 58 | Comments: 0 | Views: 233
of x
Download PDF   Embed   Report

Comments

Content


Cloud Computing



The George Washington University
Washington DC

Tarek El-Ghazawi
Observations- What led
us to Cloud Computing!
3 Tarek El-Ghazawi, GWU
Evolution of Internet Computing
P
u
b
l
i
s
h

I
n
f
o
r
m

I
n
t
e
r
a
c
t

I
n
t
e
g
r
a
t
e

T
r
a
n
s
a
c
t

D
i
s
c
o
v
e
r

(
i
n
t
e
l
l
i
g
e
n
c
e
)

A
u
t
o
m
a
t
e

(
d
i
s
c
o
v
e
r
y
)

time
scale
S
o
c
i
a
l

m
e
d
i
a

a
n
d

n
e
t
w
o
r
k
i
n
g

Semantic
discovery
Data-intensive
HPC, cloud web
deep web
D
a
t
a

m
a
r
k
e
t
p
l
a
c
e

a
n
d

a
n
a
l
y
t
i
c
s

Wipro Chennai 2011
3
4 Tarek El-Ghazawi, GWU
Top Ten Largest Databases
0
1000
2000
3000
4000
5000
6000
7000
LOC CIA Amazon YOUTube ChoicePt Sprint Google AT&T NERSC Climate
Top ten largest databases (2007)

Terabytes
Ref: http://www.focus.com/fyi/operations/10-largest-databases-in-the-world/
4
5 Tarek El-Ghazawi, GWU
Data Center
A data center is a facility used for housing a large
amount of computers that
store and serve vast amounts of data.
Source: Google
Hundreds of thousands of computers
Peta- and Exa-scale datasets
6 Tarek El-Ghazawi, GWU
Data Center vs Supercomputers
 Scale
Blue Waters = 40K 8-core “ servers”
Road Runner = 13K cell + 6K AMD servers
MS Chicago Data Center = 50 containers = 100K 8-core
servers.
 Network Architecture
Supercomputers: CLOS “ Fat Tree” infiniband
 Low latency – high bandwidth
 protocols
Data Center: IP based
 Optimized for Internet Access
 Data Storage
Supers: separate data farm
 GPFS or other parallel file system
DCs: use disk on node +
memcache
Fat tree network
Standard Data Center Network
7 Tarek El-Ghazawi, GWU
GWU HPCL Facility
MY Own HPCL Data
Center
8 Tarek El-Ghazawi, GWU

9 Tarek El-Ghazawi, GWU

10 Tarek El-Ghazawi, GWU
11 Tarek El-Ghazawi, GWU
GEORGE- GW CRAY
XE6/XK7
 ~2000 Processor Core
 50+ TF
 Based on
16 Core AMD Bulldozer chips
12-core 64-bit AMD Opteron 6100
Series processors
Kepler GPGPUs
 64 GB registered ECC DDR3
SDRAM per compute node
 1 Gemini routing and
communications ASIC per two
compute nodes
12 Tarek El-Ghazawi, GWU

13 Tarek El-Ghazawi, GWU
A Tour in the Google Data Center
 http://www.youtube.com/watch?v=zRwPSFpLX8I

11 football fields
Google Data Center
The Dalles, Oregon
Source: NY Times

Microsoft Data Center @ San Antonio, Texas
15 Tarek El-Ghazawi, GWU
What matters is on the inside







*EPA report to Congress on Server and Data Center Energy Efficiency, August 2007

SEAS Seminar Series
L Barroso and U Holzle, The Datacenter as a Computer, 2009
16 Tarek El-Ghazawi, GWU
Number Games
Microsoft San Antonio Data Center:
$550 million in 2008
475,000 square feet, or about 11 acres
A large supply of nuclear and wind energy
 Future plans include solar panels
8 million gallons of recycled waste water/month

Energy used by US data centers
61 billion kWh in 2006
$4.5 billion
1.5% of US electricity consumption
More than that of all TVs in US
Equal to that of 5.8 million households
Doubled from 2000 to 2006
Expected to double again by 2011

*EPA report to Congress on Server and Data Center Energy Efficiency, August 2007
17 Tarek El-Ghazawi, GWU
Forbes Predictions 2011
Cloud Adopters Embrace Cloud For Both
Innovation and Legacy Optimization
Replace most new procurement with cloud
strategies.
Start with private clouds as a stepping stone to
public clouds.
Get real about security. Move to private clouds
as a back up to public clouds.
The Bottom Line: Cloud Adoption Provide A
Path To The Next Generation Enterprise
18 Tarek El-Ghazawi, GWU
Internet 2: 100 Gigabit Network Infrastructure
18
19 Tarek El-Ghazawi, GWU
Perils of Corporate Computing
Own information systems 
However
Capital investment 
Heavy fixed costs 
Redundant expenditures 
High energy cost, low CPU utilization 
Dealing with unreliable hardware 
High-levels of overcapacity (Technology and Labor)

NOT SUSTAINABLE
20 Tarek El-Ghazawi, GWU
Google: CPU Utilization

Activity profile of a sample of 5,000 Google Servers over a period of 6 months
21 Tarek El-Ghazawi, GWU
Google: Energy Overhead

22 Tarek El-Ghazawi, GWU
Google: Service Disruptions

23 Tarek El-Ghazawi, GWU
Utility Computing
“ Computing may someday be organized as a
public utility, just as the telephone system is
organized as a public utility”
John McCarthy, 1961
24 Tarek El-Ghazawi, GWU
Why Utility Computing Now
Large data stores
Fiber networks
Commodity computing
Multicore machines
+
Huge data sets
Utilization/Energy
Shared people
> Utility Computing
25 Tarek El-Ghazawi, GWU
Utility Computing
Let economy of scale prevail
Outsource all the trouble to someone
else
The utility provider will share the
overhead costs among many customers,
amortizing the costs
You only pay for:
 the amortized overhead
 Your real CPU / Storage / Bandwidth
usage
26 Tarek El-Ghazawi, GWU
Data Intensive Computing
Data collection too large to transmit
economically over Internet --- Petabyte data
collections
Computation produces small data output
containing a high density of information
Implemented in Clouds
Easy to write programs, fast turn around.
MapReduce.
Map(k1, v1) -> list (k2, v2)
Reduce(k2,list(v2)) -> list(v3)
Hadoop, PIG, HDFS, Hbase
Sawzall, Google File System, BigTable
27 Tarek El-Ghazawi, GWU
Cloud Challenges
 Alignment with the needs of the business / user /
non-computer specialists / community and society
 Need to address the scalability issue: large scale
data, high performance computing, automation,
response time, rapid prototyping, and rapid time to
production
 Need to effectively address (i) ever shortening
cycle of obsolescence, (ii) heterogeneity and (iii)
rapid changes in requirements
 Transform data from diverse sources into
intelligence and deliver intelligence to right
people/user/systems
 What about providing all this in a cost-effective
manner?




6/23/2010 Wipro Chennai 2011 27
28 Tarek El-Ghazawi, GWU
Cloud Computing- A Formal Perspective

29 Tarek El-Ghazawi, GWU
Cloud Computing
Cloud computing – NIST definition
A model for enabling ubiquitous, convenient, on-
demand network access to a shared pool of
configurable computing resources
Networks, Servers, Storage, Applications, Services,

That can be rapidly provisioned and released with
minimal management effort or service provider
interaction. This cloud model is composed of five
essential characteristics, three service models, and four
deployment models.

Preston A. Cox: Mobile cloud computing: Devices, trends, issues, and the enabling
technologies
29
30 Tarek El-Ghazawi, GWU
The Five Essential Cloud Characteristics
 On-demand self-service
A consumer can unilaterally provision computing
capabilities as needed
 Broad network access
Capabilities are available over the network
 Resource pooling
Resources are pooled to serve multiple consumers
 Rapid elasticity
Capabilities can be elastically provisioned and released
 Measured service
Resource usage can be monitored and quantified
31 Tarek El-Ghazawi, GWU
The Three “ BASIC” Cloud Service Model

 SaaS- Software as a Service
- Rents software on a subscription basis
- Service includes software, hardware and support
- Users access the service through authorized device
- Suitable for a company to outsource hosting of apps
 PaaS – Platform as a Service
- Vendor offers development environment to application
developers
- Provide develops toolkits, building blocks, payment hooks
 IaaS – Infrastructure as a Service
- Processing power and storage service
32 Tarek El-Ghazawi, GWU
Cloud Computing – Services cont.

33 Tarek El-Ghazawi, GWU
The Four Cloud Deployment Models
 Private cloud: The cloud infrastructure is provisioned
for exclusive use by a single organization
 Community cloud: The cloud infrastructure is
provisioned for exclusive use by a specific community
of consumers from organizations that have shared
concerns (e.g., mission, security requirements, policy,
and compliance considerations)
 Public cloud: The cloud infrastructure is provisioned
for open use by the general public.
 Hybrid cloud: The cloud infrastructure is a
composition of two or more distinct cloud
infrastructures (private, community, or public)
34 Tarek El-Ghazawi, GWU
Software Stack
Mobile (Android), Thin client (Zonbu) Thick
client (Google Chrome)
Identity, Integration Payments, Mapping,
Search, Video Games, Chat
Peer-to-peer (Bittorrent), Web app (twitter),
SaaS (Google Apps, SAP)
Java Google Web Toolkit, Django, Ruby on
Rails, .NET
S3, Nirvanix, Rackspace Cloud Files,
Savvis,
Full virtualization (GoGrid), Management
(RightScale), Compute (EC2), Platform
(Force.com)

35 Tarek El-Ghazawi, GWU
NIST: Interactions between Actors in Cloud
Computing
35
Cloud
Consumer
Cloud Provider
Cloud Broker
Cloud Auditor

The communication path between a cloud provider & a cloud consumer
The communication paths for a cloud auditor to collect auditing information
The communication paths for a cloud broker to provide service to a cloud
consumer


36 Tarek El-Ghazawi, GWU
Windows Azure
 Enterprise-level on-demand capacity builder
 Fabric of cycles and storage available on-request for a
cost
 You have to use Azure API to work with the
infrastructure offered by Microsoft
 Significant features: web role, worker role , blob
storage, table and drive-storage
6/23/2010 Wipro Chennai 2011 36
37 Tarek El-Ghazawi, GWU
Google App Engine
 This is more a web interface for a development
environment that offers a one stop facility for design,
development and deployment Java and Python-based
applications in Java, Go and Python.
 Google offers the same reliability, availability and
scalability at par with Google’s own applications
 Interface is software programming based
 Comprehensive programming platform irrespective of
the size (small or large)
 Signature features: templates and appspot, excellent
monitoring and management console
6/23/2010 Wipro Chennai 2011 37
38 Tarek El-Ghazawi, GWU
Amazon EC2
 Amazon EC2 is one large complex web service.
 EC2 provided an API for instantiating computing
instances with any of the operating systems
supported.
 It can facilitate computations through Amazon Machine
Images (AMIs) for various other models.
 Signature features: S3, Cloud Management Console,
MapReduce Cloud, Amazon Machine Image (AMI)
 Excellent distribution, load balancing, cloud
monitoring tools
6/23/2010 Wipro Chennai 2011 38
39 Tarek El-Ghazawi, GWU
39
Amazon Elastic Compute Cloud
“ a web service that provides resizable
compute capacity in the cloud. It is designed
to make web-scale computing easier for
developers. “

Create Amazon Machine Image (AMI)
Upload the AMI into Amazon S3
Use Amazon EC2 web service to manage
Pay as you go
40 Tarek El-Ghazawi, GWU
Amazon Simple Storage Service
 “ storage for the Internet. It is designed to make web-
scale computing easier for developers.”

 Write, read, and delete objects
 Unlimited objects
 Authorization mechanisms
 REST and SOAP interfaces
 HTTP/BitTorrent protocol
40
41 Tarek El-Ghazawi, GWU
Amazon Pricing
 Compute
$0.10 - Small Instance (Default) 1.7 GB of memory, 1 EC2 Compute
Unit (1 virtual core with 1 EC2 Compute Unit), 160 GB of instance
storage, 32-bit platform
$0.40 - Large Instance
$0.80 - Extra Large Instance

 Data Transfer
$0.100 per GB - all data transfer in $0.170 per GB - first 10 TB / month
data transfer out
$0.130 per GB - next 40 TB / month data transfer out
$0.110 per GB - next 100 TB / month data transfer out
$0.100 per GB - data transfer out / month over 150 TB
 Looks inexpensive, but really?

41
42 Tarek El-Ghazawi, GWU
Putting Numbers Together
EC2
1K instance hours, 1TB data in & out = $370
10K instance hours, 1TB data in & out = $1,270
100K instance hours, 1TB data in & out = $100,270

S3
10TB storage, 100GB data in &out = 1,527.00
100TB storage, 1 TB data in &out = 15,270.00
1PB storage, 10 TB data in &out = 152,700.00

42
43 Tarek El-Ghazawi, GWU
Downtime
 “ 7.1. Downtime and Service Suspensions. In addition
to our rights to terminate or suspend Services to you
as described in Section 3 above, you acknowledge
that: (i) your access to and use of the Services may be
suspended for the duration of any unanticipated or
unscheduled downtime or unavailability of any portion
or all of the Services for any reason, including as a
result of power outages, system failures or other
interruptions; and (ii) we shall also be entitled, without
any liability to you, to suspend access to any portion
or all of the Services at any time, on a Service-wide
basis …”
43
44 Tarek El-Ghazawi, GWU
Security
 “ 7.2. Security. We strive to keep Your Content secure,
but cannot guarantee that we will be successful at
doing so, given the nature of the Internet. Accordingly,
without limitation to Section 4.3 above and Section
11.5 below, you acknowledge that you bear sole
responsibility for adequate security, protection and
backup of Your Content. We strongly encourage you,
where available and appropriate, to use encryption
technology to protect Your Content from unauthorized
access and to routinely archive Your Content. We will
have no liability to you for any unauthorized access or
use, corruption, deletion, destruction or loss of any of
Your Content.”
44
45 Tarek El-Ghazawi, GWU
Amazon S3 SLA
45
46 Tarek El-Ghazawi, GWU
 Open Cloud Computing Interface – Infrastructure
 EC2 API
 Simple Storage Service (S3) API
 Windows Azure Storage Service REST APIs
 Windows Azure Service Management REST APIs
 Deltacloud API
 Rackspace Cloud Servers API
 Rackspace Cloud Files API
 Cloud Data Management Interface
 vCloud API
 GlobusOnline REST API
46
Cloud Interoperability Standards
47 Tarek El-Ghazawi, GWU
Challenges
 Alignment with the needs of the business / user / non-
computer specialists / community and society
 Need to address the scalability issue: large scale data,
high performance computing, automation, response
time, rapid prototyping, and rapid time to production
 Need to effectively address (i) ever shortening cycle of
obsolescence, (ii) heterogeneity and (iii) rapid changes in
requirements
 Transform data from diverse sources into intelligence
and deliver intelligence to right people/user/systems
 What about providing all this in a cost-effective manner?




6/23/2010 Wipro Chennai 2011 47
48 Tarek El-Ghazawi, GWU
Conclusions
 Economics and technology advances side by side
drive our IT transformation
 One such transformation is utility computing and the
cloud which is driven by
Necessity of having excess capacity to deal with
processing
Rising complexity of keeping and maintaining IT
capabilities in house
The explosion of data and its distribution across the
world

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close