Dr.

AMBEDKAR INSTITUTE OF TECHNOLOGY

OUTER RING ROAD,MALLATHALLI,BENGALURU
DEPARTMENT OF COMPUTER SCIECNE AND ENGINEERING

Cloud Computing –Kai Hwang

Unit 1
Scalable Computing Over the Internet
System Models for Distributed and Cloud Computing
Software Environments for Distributed Systems and Clouds: Performance, Security and
Energy Efficiency

https://drait.edu.in 1
 Cloud computing is the on-demand availability of computer systems,
especially data storage (cloud storage) and computing power, without
direct active management by the user.

 Large clouds often have functions distributed over multiple locations,

each of which is a Data center.

 Cloud computing relies on sharing of resources “pay as you go"

model.
 Unit 1
Scalable Computing Over the Internet: The Age of Internet Computing, Scalable
Computing Trends and New Paradigms, Virtual Machines and Virtualization Middleware,
Data Center Virtualization for Cloud Computing.

System Models for Distributed and Cloud Computing: Clusters of Cooperative

Computers, Grid Computing Infrastructures, Peer-to-Peer Network Families, Cloud
Computing over the Internet,

Software Environments for Distributed Systems and Clouds: Service-Oriented

Architecture (SOA), Performance, Security and Energy Efficiency: Performance Metrics
and Scalability Analysis, Fault Tolerance and System Availability, Network Threats and
Data Integrity, Energy Efficiency in Distributed Computing

5
1.1.1 The Age of Internet Computing
1. Linpack Benchmark for high-performance computing (HPC) applications is
obselute for measuring system performance.
2.The emergence of computing clouds instead demands high-throughput
computing (HTC) systems built with parallel and distributed computing
technologies.
3.Need to upgrade data centers using fast servers, storage systems, and high-
bandwidth networks.
4.From 1970 to 1990, widespread with VLSI microprocessors.
5.From 1980 to 2000, massive numbers of portable both wired and wireless
applications.
6.Since 1990, HPC and HTC clusters, grids, or Internet clouds

6
 Figure 1.1 evolution of HPC and HTC systems.
HPC -Supercomputers (massively parallel processors) by clusters of cooperative computers to share computing
resources.
The cluster is often a collection of homogeneous compute nodes that are physically connected in close range to one
another.
HTC -peer-to-peer (P2P) distributed file sharing and content delivery applications.
Service-Oriented Architecture (SOA) : Applications make use of services available in the network.
High-Performance Computing-HPC

1.Raw speed performance.

2.Gflops(floating-point speed) Pflops -scientific, engineering, and

manufacturing communities.
High-Throughput Computing

1. HPC paradigm to an HTC paradigm.

2. More attention to high-flux computing.
3. High-flux computing --- Internet searches and web services by
millions or more users simultaneously.
4. cost, energy savings, security, and reliability for enterprise
computing.
Three New Computing Paradigms

 SOA, Web 2.0 services become available.

The maturity of radio-frequency identification (RFID),
Global Positioning System (GPS), and sensor technologies
leads the development of the Internet of Things (IoT).
 Computing Paradigm Distinctions

 Centralized computing:
 Parallel computing:
 Distributed computing
 Cloud computing
 Ubiquitous computing:
 The Internet of Things (IoT)
 Internet computing .
 Design objectives
1. Efficiency:
 resources utilization for massive parallelism
 efficiency is job throughput, data access, storage, and power efficiency.
2. Dependability:
 The reliability
 self-management with Quality of Service (QoS) assurance, even under failure
conditions.
3. Adaptation: support billions of job requests massive data sets and virtualized
cloud resources under various workload and service models.
4.Flexibility: Application deployment both HPC (science and engineering) and HTC (business)
applications
 1.1.2 Scalable Computing Trends and New Paradigms
 resource distribution and concurrency or high degree of parallelism
(DoP).
 Bit-level parallelism (BLP) converts bit-serial processing to word-level processing
gradually.
 Instruction-level parallelism (ILP), in which the processor executes multiple
instructions simultaneously rather than only one instruction at a time.
 ILP requires branch prediction, dynamic scheduling, speculation, and compiler
support to work efficiently.
 Data-level parallelism (DLP) through SIMD (single instruction, multiple data).
 DLP --hardware support and compiler assistance to work properly.
Innovative Applications
Trend toward Utility Computing

FIGURE 1.2:The vision of computer utilities in modern distributed computing systems

First, Reliability and scalability are two major design objectives.

Second, autonomic operations that can be self-organized to support dynamic discovery.
Finally, with QoS and SLAs (service-level agreements)
 The Hype Cycle of New Technologies

1.New and emerging computing and information technology may go through a hype cycle
 The Internet of Things and Cyber-Physical Systems
Interconnection, tools, devices, or computers.
RFID or a related sensor or electronic technology -
GPS.
IPv6 protocol, 2128 IP addresses.
surrounded by 1,000 to 5,000 objects.
Designed to track 100 trillion static or moving
objects simultaneously.
Demands universal addressability
 All objects and devices are instrumented, interconnected,
and interacted with each other intelligently.
H2H (human-to-human)
 H2T (human- to thing)
T2T (thing-to-thing). things are PCs and mobile phones.
with low cost.
Cyber-Physical Systems
 Cyber-Physical Systems
CPS  interaction between computational processes and
the physical world.
“cyber” (heterogeneous, asynchronous) with “physical”
(concurrent and information-dense) objects.
A CPS merges the “3C” computation, communication,
and control.
The IoT ----- various networking connections among
physical objects
CPS --- virtual reality (VR) applications in the physical
world.
https://www.youtube.com/watch?v=eXsNX_2AzM8
 Virtual Machines and Virtualization Middleware
Virtualization relies on software to simulate hardware
functionality and create a virtual computer system. scale and
greater efficiency.
A virtual machine (VM)
Virtualization/emulation of computer system.
Virtual machines are based on computer architectures and
provide functionality of a physical computer.
specialized hardware, software, or a combination.
 Virtual Machines and Virtualization Middleware

1. A hypervisor, also known as a virtual machine monitor or VMM,

is software that creates and runs virtual machines (VMs).
2. A hypervisor allows one host computer to support multiple guest
VMs by virtually sharing its resources, such as memory and
processing.
 Virtual Machines and Virtualization Middleware

1.A conventional computer has a single OS image.

2.Virtual machines (VMs)reduces underutilized resources,
application inflexibility, software manageability, and
security concerns in existing physical machines.
3.virtualization of processors, memory, and I/O facilities
dynamically.
Part a – Physical Machine
 The host machine is equipped with the physical
hardware.
Part b – Native VM
The VM is built with virtual resources managed by a
guest OS to run a specific application.
Between the VMs and the host platform, deploy a
middleware layer called a virtual machine monitor
(VMM).
VMM runs Privileged mode
 bare-metal VM, because the hypervisor handles
the bare hardware (CPU, memory, and I/O) direct
Eg: the hardware has x-86 architecture running the Windows
system
 Part c – Hosted VM
Here the VMM runs in non-privileged mode
 The host OS need not be modified
Part d – Dual – mode VM
The VM can also be implemented with a dual
mode Part of the VMM runs at the user level
Part d – Dual – mode VM
and another part runs at the supervisor level.
The VM can also be implemented with a dual mode
The
Part host
of the VMMOS may
runshave touser
at the be modified
level andto some part runs at the
another
extent. level.
supervisor
Inthis
Multiple VMs
case, the host can
OS may be have
ported
to betomodified
a given to some extent. Multiple VMs
can hardware
be ported tosystem tohardware
a given support the virtualization
system to support the virtualization process.
process.
VM Primitive Operations

 The VMM provides the VM abstraction to the guest OS.

 Low-level VMM operations are indicated Figure 1.13 : Multiplexing,
Suspension, Provision (resume) and Life Migration
First, the VMs can be multiplexed between hardware machines
Second, a VM can be suspended and stored in stable storage
Third, a suspended VM can be resumed or provisioned to a new
hardware platform
Finally, a VM can be migrated from one hardware platform to another

• VM approach will significantly enhance the utilization of server

resources.
• Multiple server functions can be consolidated on the same
hardware platform to achieve higher system efficiency
 Data Center Virtualization for Cloud Computing

1.Data Center Growth and Cost Breakdown

Cloud architecture - Hardware and network devices.

Low-cost terabyte disks and Gigabit Ethernet are used to
build data centers.
 Data center  performance/price ratio over speed
performance alone.
 Storage and energy efficiency are more important than
shear speed performance
 Data Center Virtualization for Cloud Computing
2.Low-Cost Design Philosophy

 High-end switches or routers

 High-bandwidth networks.
 Commodity switches and networks are more suitable in data
centers because of the fixed budget.
 x86 servers is more desired over expensive mainframes.
 The software layer handles network traffic balancing, fault
tolerance, and expandability.
 Data Center Virtualization for Cloud Computing

3.Convergence of Technologies
 Hardware virtualization and multi-core chips - enable the
existence of dynamic configurations in the cloud
 utility and grid computing
 SOA, Web 2.0, and data center automation
 System Models for Distributed and Cloud Computing

Distributed and Cloud computing systems :

• Built over a large number of autonomous computer nodes.

• Interconnected by SANs, LANs, or WANs in a hierarchical manner.
• LAN switches connect hundreds of machines as a working cluster.
• WAN connect many local clusters to form a very large cluster of
clusters.
Massive systems are considered highly scalable, and can reach web-scale
connectivity, either physically or logically.

Massive systems are classified into four groups:

 Clusters
 P2P networks
 Computing grids
 Internet clouds over huge data centers
These four system classes may involve hundreds, thousands, or even
millions of computers as participating nodes
Clusters of Cooperative Computers
• Cluster Architecture
 Cluster Architecture
 A typical server cluster built around a low-latency, high
bandwidth interconnection network.
 Network can be:
• a simple SAN (e.g., Myrinet) • a LAN (e.g., Ethernet)
 larger cluster - with multiple levels of Gigabit Ethernet,
Myrinet,switches.
 The cluster is connected to the Internet via a virtual private
network (VPN) gateway.
 The gateway IP address locates the cluster
Myrinet at Barcilona data center
Single System Image (SSI):

 Ideal cluster--merge multiple system images into a single-system

image.
 A cluster operating system or some middleware- to support SSI like
sharing of CPUs, memory, and I/O across all cluster nodes.
 SSI illusion created by software or hardware that presents a
collection of resources as one integrated, powerful resource.
 SSI makes the cluster appear like a single machine to the user.
 A cluster with multiple system images is nothing but a collection of
independent computers.
 Hardware, Software, and Middleware Support:
1. Hardware: PCs, workstations, servers, or SMP(systematic
multiprocessing)
2. Software: PVM(parallel virtual machine) or MPI(Message passing
interface)
3. The computer nodes are interconnected by a high-bandwidth
network (such as Gigabit Ethernet, Myrinet).
4.Middleware:Type of software to provide software applications beyond
those available from the operating system.
 Major Cluster Design Issues
1. A cluster-wide OS for complete resource sharing.
2. Middleware or OS extensions
3. Middleware- to achieve cooperative computing.
4. Middleware- achieve high performance.
5. Scalable performance, efficient message passing, high
system availability, seamless fault tolerance, and cluster-
wide job management.
 Grid Computing Infrastructures
Grid computing ---- allow close interaction on distant
computers simultaneously.
 Computational Grids
 A computing grid that couples computers, software/middleware,
special instruments, and people and sensors together.
 The grid is across LAN, WAN, or Internet backbone networks at a
regional, national, or global scale.
 Enterprises or organizations present grids as integrated computing
resources.
 Personal computers, laptops, and PDAs can be used as access devices to a grid
system.
Figure 1.16 shows an example computational grid built over multiple resource sites owned by different organizations. The
resource sites offer complementary computing resources, including workstations, large servers, a mesh of processors, and
Linux clusters to satisfy a chain of computational needs.
 Grid Families Requirement

 Models, software/middleware support, network

protocols, and hardware infrastructures.
 IBM, Microsoft, Sun, HP, Dell, Cisco, EMC, Platform
Computing,
 New grid service providers (GSPs) and new grid
applications
 Peer-to-Peer Network Families

 Client machines (PCs and workstations) are connected to a

central server for compute, e-mail, file access, and database
applications.

 The P2P architecture -a distributed model of networked

systems.

 First, a P2P network is client-oriented instead of server-

oriented.
 Overlay Networks
Two types of overlay networks:
unstructured and structured
•An unstructured overlay network is characterized by a random graph.
•There is no fixed route to send messages or files among the nodes.
•Often, flooding is applied to send a query to all nodes in an unstructured
overlay, thus resulting in heavy network traffic and nondeterministic search
results.

•Structured overlay networks follow certain connectivity topology and rules

for inserting and removing nodes (peer IDs) from the overlay graph.
•Routing mechanisms are developed to take advantage of the structured
overlays.
 P2P Application Families

Gnutella, Napster, and BitTorrent

Collaboration P2P networks include MSN(Microsoft) or
Skype chatting, instant messaging, and collaborative design,
among others.
The third family is for distributed P2P. For example,
SETI@home(Scietific experiment for Extraterrestrial
Intelligence) provides 25 Tflops of 3 million Internet host
machines.
Other P2P platforms, such as JXTA(Juxtapose), .NET, and
FightingAID@home(Fighting pollution attacks).
 P2P Computing Challenges
 Heterogeneity problems: in hardware, software, and network
requirements.
 incompatibility exists between software and the OS;
 scalability as the workload increases-with Bandwidth
optimization.
 Data locality, network proximity, and interoperability are three
design objectives in distributed P2P applications.
 P2P performance is affected by routing efficiency and self-
organization by participating peers.

 P2P Computing Challenges

 Fault tolerance, failure management, and load balancing are

other important issues in using overlay networks.
 Lack of trust
 Security, privacy, and copyright violations
 In a P2P network, all clients provide resources including
computing power, storage space, and I/O bandwidth.
Disadvantages of P2P networks:

 Because the system is not centralized, managing it is difficult.

 The system lacks security. Anyone can log on to the system and
cause damage or abuse.

P2P networks are reliable for a small number of peer nodes.

useful for applications that require a low level of security and have no
concern for data sensitivity.
 Cloud Computing over the Internet
Definition of Cloud Computing by IBM:

Pool of virtualized computer resources.

Different workloads batch-style backend jobs and interactive and
user-facing applications
 redundant, self-recovering, highly scalable programming models
that allow workloads to recover from many unavoidable
hardware/software failures.
Finally, monitor resource use in real time to enable rebalancing of
allocations when needed.
Internet Clouds
 virtualized platform
with elastic
resources on
demand .
 satisfy many user
applications
simultaneously.
 designed to be
secure, trustworthy,
and dependable.
 The Cloud Landscape

1.Infrastructure as a Service (IaaS):

 Infrastructures demanded by users—servers, storage, networks,
and the data center fabric.
 The user can deploy and run on multiple VMs running guest OS’s
on specific applications.
 The user does not manage or control, but can specify when to
request and release the needed resources.
 The Cloud Landscape
2.Platform as a Service (PaaS) :
 user to deploy user-built applications
 middleware, databases, development tools, and some runtime
support such as Web 2.0 and Java.
 both hardware and software integrated with specific programming
interfaces.
 The provider supplies the API and software tools (e.g., Java,
Python, Web 2.0, .NET).
 The user is freed from managing the cloud infrastructure.
 The Cloud Landscape
3.Software as a Service (SaaS):
 Browser-initiated application software over thousands of paid cloud
customers.
 Business processes, industry applications, consumer relationship
management (CRM), enterprise resources planning (ERP), human
resources (HR), and collaborative applications.
 On the customer side, there is no upfront investment in servers or
software licensing.
 On the provider side, costs are rather low, compared with
conventional hosting of user applications.
 Eight reasons to adapt the cloud for upgraded Internet applications and web
services:

1.Desired location with protected space and higher energy

efficiency
2.Sharing of peak-load capacity among a large pool of users,
improving overall utilization
3.Separation of infrastructure maintenance duties from
domain-specific application development
4. Significant reduction in cloud computing cost, compared with
traditional computing paradigms
 The following list highlights eight reasons to adapt the cloud for upgraded
Internet applications and web services:

5. Cloud computing programming and application development

6. Service and data discovery and content/service distribution
7. Privacy, security, copyright, and reliability issues
8. Service agreements, business models, and pricing policies
SOFTWARE ENVIRONMENTS FOR DISTRIBUTED SYSTEMS
AND CLOUDS
Service-Oriented Architecture (SOA)
Layered Architecture for Web Services and Grids
 The entity interfaces correspond to the Web
Services Description Language (WSDL), Java
method, and CORBA interface definition language
(IDL) specifications in distributed systems.
 These interfaces are linked with customized, high-
level communication systems: SOAP(Simple object
access protocol),REST, RMI, and IIOP (Internet Inter-ORB Protocol)
 Communication systems -(Remote Procedure Call -
RPC), fault recovery, and specialized routing.
Fault tolerance-Web Services Reliable Messaging (WSRM)
framework mimic the OSI layer capability modified to
match the different abstractions at the entity levels.
Security is a critical capability uses Internet Protocol
Security (IPsec) and secure sockets in the OSI layers.
 Web Services and Tools
 Specify complete aspects of the web service and its environment.
Simple Object Access Protocol (SOAP).
SOAP requests are easy to generate and process responses.
a request by a an XML document.

SOAP client sends the XML document to a SOAP server.

When the server receives the SOAP message, it sends the message as
a service invocation to the requested server-side application.
A response- Requested parameters, return values and data for the
client is returned first to the SOAP request handler and then to the
requesting client.
REST
 Adopts simplicity as the universal principle
 REST has minimal information in the header, and the message body
carries all the needed information.
 REST can use XML schemas but not those that are part of SOAP;
 The Evolution of SOA

 SOA applies to building grids, clouds, grids of clouds, clouds of grids, clouds of
clouds.
 A large number of sensors provide data-collection services as SS (sensor
service).
 A sensor can be a ZigBee device, a Bluetooth device, a WiFi access point, a
personal computer, a GPA(generalized power allocation), or a wireless phone,
among other things.
 Raw data is collected by sensor services.
 All the SS devices interact with large or small computers, many forms of grids,
databases, the compute cloud, the storage cloud, the filter cloud, the discovery
cloud, and so on.
 The Evolution of SOA

Filter services ( fs) are used to eliminate unwanted raw data, in

order to respond to specific requests from the web, the grid, or web
services.
A collection of filter services forms a filter cloud.
SOA aims to search for, or sort out, the useful data from the massive
amounts of raw data items.
Processing this data will generate useful information, and
subsequently, the knowledge for our daily use.
wisdom or intelligence is sorted out of large knowledge bases.
Finally,intelligent decisions based on both biological and machine
wisdom
The Evolution of SOA
 Grids versus Clouds
Grid - static resources, cloud - elastic resources.
The differences between grids and clouds are limited only in
dynamic resource allocation based on virtualization and
autonomic computing.
cloud of clouds, a grid of clouds, or a cloud of grids, or inter-
clouds as a basic SOA architecture.
 Performance Metrics and Scalability Analysis
1. Performance Metrics
System throughput in MIPS, Tflops(Tera floating-point operations per
second), or TPS (transactions per second). Other measures include job
response time and network latency.
An interconnection network that has low latency and high bandwidth is
preferred.
 System overhead - OS boot time, compile time, I/O data rate, and the
runtime support system used.
QoS for Internet and web services; system availability and
dependability; and security resilience for system defense against
network attacks.
2.Dimensions of Scalability

 Size scalability: Machine size. “size” adding processors, cache, memory,

storage, or I/O channels by counting the number of processors installed.

 Software scalability :Upgrades in the OS or compilers, adding

mathematical and engineering libraries, porting new application software,
and installing more user-friendly programming environments.

 Application scalability :Refers to matching problem size scalability with

machine size scalability.
 Problem size affects the size of the dataset or the workload increase.
 Instead of increasing machine size, users can enlarge the problem size to
enhance system efficiency or cost-effectiveness.
2.Dimensions of Scalability

Technology scalability : adapt to technologies, system design time, space,

and Heterogeneity
(1) Time refers to generation scalability. When changing to new-
generation processors motherboard, power supply, packaging and
cooling so forth.
(2) Space is related to packaging and energy concerns. Technology
scalability demands harmony and portability among suppliers.
(3) Heterogeneity hardware components or software packages from
different vendors. Heterogeneity may limit the scalability.
3. Scalability versus OS Image

1.Scalable performance - higher speed by adding more processors or

servers, with memory, disk capacity, I/O channels.
2.The OS image is counted by the number of independent OS images
observed in a cluster, grid, P2P network, or the cloud.
3.SMP(symmetric multiprocessor) and NUMA(nonuniform memory
access) are included in the comparison.
4.An SMP has a single system image, which could be a single node in a
large cluster.
 Performance Metrics and Scalability Analysis
3. Scalability versus OS Image

5.NUMA machines are often made out of SMP nodes with distributed, shared
memory.
6 A NUMA machine can run with multiple operating systems, and can scale to
a few thousand processors communicating with the MPI library
 Performance Metrics and Scalability Analysis

4. Amdahl’s Law
Consider the execution of a given program on a uniprocessor
workstation with a total execution time of T minutes. Now, let’s say
the program has been parallelized or partitioned for parallel
execution on a cluster of many processing nodes. Assume that a
fraction α of the code must be executed sequentially, called the
sequential bottleneck. Therefore, (1 − α) of the code can be
compiled for parallel execution by n processors. The total execution
time of the program is calculated by α T + (1− α)T/n, where the first
term is the sequential execution time on a single processor and the
second term is the parallel execution time on n processing nodes.
Amdahl’s Law states that the speedup factor of using the n-processor
system over the use of a single processor is expressed by:

Speedup =S= T/[αT +(1− α)T/n] =1/[α +(1 −α)/n]

The maximum speedup of n is achieved only if the sequential
bottleneck α is reduced to zero or the code is fully parallelizable with
α = 0. with large cluster n → ∞, S approaches 1/α, an upper bound on
the speedup S and is independent of the cluster size n.
The sequential bottleneck is the portion of the code that cannot be parallelized.
For example, the maximum speedup achieved is 4, if α = 0.25 or 1 − α = 0.75, even if one
uses hundreds of processors.
 Performance Metrics and Scalability Analysis

5. Problem with Fixed Workload

In Amdahl’s law, we have assumed the same amount of workload for both
sequential and parallel execution of the program with a fixed problem size or
dataset. To execute a fixed workload on n processors, parallel processing may
lead to a system efficiency defined as follows:

E =S/n =1/[αn +1 −α ]
 System efficiency is low, when the cluster size is very large.
 A cluster with n = 256 nodes, extremely low efficiency
E = 1/[0.25 × 256 + 0.75] = 1.5% is observed
 Performance Metrics and Scalability Analysis
6. Gustafson’s Law
 To achieve higher efficiency when using a large cluster, consider
scaling the problem size to match the cluster capability-speedup law
proposed by John Gustafson as scaled-workload speedup.
 Let W be the workload in a given program.
 When using an n-processor system, the user scales the workload to
W′ = αW + (1 − α)nW.
 Scaled workload W′ is essentially the sequential execution time on a
single processor.
 Performance Metrics and Scalability Analysis
6. Gustafson’s Law
The parallel execution time of a scaled workload W′ on n processors is
defined by a scaled-workload speedup as follows:
S′ =W′/W =[αW + (1− α)nW]/W = α+ (1 −α)n
This speedup is known as Gustafson’s law. By fixing the parallel
execution time at level W,

E′ =S′/n= α/n+(1− α)
256-node cluster to E′ = 0.25/256 + 0.75 = 0.751.

Fixed workload, users should apply Amdahl’s law. To solve scaled problems, users should apply Gustafson’s
 Fault Tolerance and System Availability
1.System Availability
 HA (high availability) is desired in all clusters, grids, P2P networks, and cloud
systems.
 A system is highly available if it has a long mean time to failure (MTTF) and
a short mean time to repair (MTTR).
 System availability is formally defined as follows:
System Availability =MTTF/(MTTF +MTTR)
 All hardware, software, and network components may fail. Any failure that
will pull down the operation of the entire system is called a single point of
failure.
 Fault Tolerance and System Availability

1. System Availability

Design a dependable computing system with no single point of

failure.
Adding hardware redundancy, increasing component reliability, and
designing for testability will help to enhance system availability and
dependability.
Distributed system increases in size, availability decreases due to a
higher chance of failure and a difficulty in isolating the failures.
 Fault Tolerance and System Availability

1.System Availability

A grid is visualized as a hierarchical cluster of clusters. Grids

have higher availability due to the isolation of faults.
But clusters, clouds, and grids have decreasing availability as the
system increases in size.
A P2P file-sharing network operates independently with low
availability.
 Network Threats and Data Integrity

1. Threats to Systems and Networks

 Network viruses have threatened many users in widespread attacks.
 Worm Epidemic by pulling down many routers and servers, and are
responsible for the loss of billions of dollars in business, government, and
services.
Network Threats and Data Integrity
1. Threats to Systems and Networks
1. Loss of data integrity.
2.A denial of service (DoS)
3. Lack of authentication or authorization
4. Open resources such as data centers, P2P networks, and grid and
cloud infrastructures could become the next targets.
5. Malicious intrusions to these systems may destroy valuable hosts
 Network Threats and Data Integrity

2.Security Responsibilities
confidentiality, integrity, and availability
SaaS - cloud provider to perform all security functions.
IaaS - users to assume almost all security functions, but to leave
availability in the hands of the providers.
PaaS - provider to maintain data integrity and availability, but burdens
the user with confidentiality and privacy.
3. Copyright Protection

Collusive piracy is the intellectual property violations in P2P network.

Paid clients (colluders) may illegally share copyrighted content files
with unpaid clients (pirates)-online piracy.
Proactive content poisoning scheme to stop colluders and pirates
from alleged copyright infringements in P2P file sharing.
Pirates are detected with identity-based signatures and
timestamped tokens.
Stops collusive piracy from occurring without hurting legitimate P2P
clients.
 Network Threats and Data Integrity
4. System Defense Technologies
Three generations of network defense
1st generation to prevent or avoid intrusions. Using access
control policies or tokens, cryptographic systems, and so
forth.
2nd generation detected intrusions in a timely manner to
exercise remedial actions. Includes firewalls, intrusion
detection systems (IDSes), PKI services, reputation systems,
and so on.
3rd generation have more intelligent responses to
intrusions.
 Network Threats and Data Integrity

5. Data Protection Infrastructure

Security infrastructure is required to safeguard web and cloud

services.
At the user level, one needs to perform trust negotiation and
reputation aggregation over all users.
 At the application end, need to establish security precautions in
worm containment and intrusion detection
 Energy Efficiency in Distributed Computing

1 Energy Consumption of Unused Servers

To run a server farm (data center) a company has to spend a huge
amount of money for hardware, software, operational support, and
energy every year.
 Therefore, companies should thoroughly identify whether their
installed server farm , the volume of provisioned resources in terms
of utilization.
 Energy Efficiency in Distributed Computing
1 Energy Consumption of Unused Servers
 Energy Efficiency in Distributed Computing

2.Reducing Energy in Active Servers

 Use methods to decrease energy consumption in active distributed
systems with negligible influence on their performance.
 Power management issues in distributed computing platforms can be
categorized into four layers : the application layer, middleware layer,
resource layer, and network layer.
 Energy Efficiency in Distributed Computing
3 Application Layer
Design sophisticated multilevel and multi-domain energy
management applications without hurting performance.
Ensure the Relationship between performance and energy
consumption.
An application’s energy consumption depends strongly on the
number of instructions needed to execute the application and the
number of transactions with the storage unit (or memory). These
two factors (compute and storage) are correlated and they affect
completion time.
 Energy Efficiency in Distributed Computing
4 Middleware Layer
 The middleware layer acts as a bridge between the application layer and the
resource layer.
 This layer provides resource broker, communication service, task analyzer, task
scheduler, security access, reliability control, and information service capabilities.
 It is also responsible for applying energy-efficient techniques, particularly in task
scheduling.
5 Resource Layer
 Resources -computing nodes and storage units.
 Interacts with hardware devices and the operating system; Responsible for
controlling all distributed resources in distributed computing systems.
Dynamic power management (DPM) and dynamic voltage-
frequency scaling (DVFS) are two popular method
 Energy Efficiency in Distributed Computing
6. Network Layer
• The models should represent the networks comprehensively
understanding of interactions among time, space, and energy.
• New, energy-efficient routing algorithms need to be developed and
also against network attacks.
7 DVFS Method for Energy Efficiency
 It exploits the slack time (idle time) of intertask relationship.
 The slack time associated with a task is utilized to execute the task in a
lower voltage frequency.