9/1/2024

SYSTEM MODELS FOR DISTRIBUTED AND CLOUD

COMPUTING
• Distributed and cloud computing systems are built over a
UNIT I large number of autonomous computer nodes.
CLOUD ARCHITECTURE MODELS AND • Machines are interconnected by SAN(Storage Area
INFRASTRUCTURE Network)s, LANs, or WANs in a hierarchical manner.
• A few LAN switches can easily connect hundreds of
machines as a working cluster.
• Massive systems are considered highly scalable
• Massive systems are classified into four groups: clusters,
P2P networks, computing grids, and Internet clouds

Clusters of Cooperative Computers

• Cluster consists of interconnected stand-alone
computers which work cooperatively as a single
integrated computing resource
Architecture
• To build a larger cluster with more nodes, the
network can be built with multiple levels of
Gigabit Ethernet, Myrinet, or InfiniBand
switches
• using a SAN, LAN, or WAN, one can build
scalable clusters with an increasing number of
nodes.

1
 9/1/2024

A cluster of server connected by high • Cluster is connected to the Internet via a virtual
bandwidth with I/O devices private network (VPN) gateway.
• The gateway IP address locates the cluster.
• Most clusters have loosely coupled node
computers.
• All resources of a server node are managed by
their own OS.
• Most clusters have multiple system images as a
result of having many autonomous nodes
under different OS control.

Single-System Image Hardware, Software, and Middleware Support

• An ideal cluster should merge multiple system • Clusters exploring massive parallelism are
images into a single-system image (SSI). commonly known as MPPs
• Designers desire a cluster operating system or • All HPC(High Performance Computing) clusters in
some middleware to support SSI at various the Top 500 list are also MPPs
levels
• The building blocks are computer nodes (PCs,
• An SSI is an illusion created by software or workstations, servers, or SMP), special
hardware that presents a collection of communication software such as PVM(parallel
resources as one integrated, powerful resource virtual machine) or MPI(message passing
interface), and a network interface card in each
computer node.
• Most clusters run under the Linux OS.

2
 9/1/2024

• Special cluster middleware supports are needed to

create SSI or high availability (HA).
• Both sequential and parallel applications can run on
the cluster
• Many SSI features are expensive or difficult to achieve
• Instead of achieving SSI, many clusters are loosely
coupled machines.
Cluster Design Issues
• A cluster-wide OS for complete resource sharing is
not available yet
• Middleware or OS extensions were developed at the
user space to achieve SSI

Grid Computing Infrastructures Computational Grids

• Computing grid offers an infrastructure that
• Users have experienced a growth path from
couples computers, software/middleware,
Internet to web and grid computing services.
special instruments, and people and sensors
• Internet services such as the Telnet command together.
enables a local computer to connect to a
• The grid is constructed across LAN, WAN, or
remote computer.
Internet backbone networks
• A web service such as HTTP enables remote
• They can be viewed as virtual platforms to
access of remote web pages.
support virtual organizations
• Grid computing is envisioned to allow close
• The computers used in a grid are primarily
interaction among applications running on
workstations, servers, clusters, and
distant computers simultaneously.
supercomputers

3
 9/1/2024

At the server end, the grid is a network. Grid Families

At the client end wired or wireless terminal devices • Grid technology demands new distributed computing
Eg: TeraGrid in US, EGEE in Europe, and ChinaGrid in models, software/middleware support, network
protocols, and hardware infrastructures.
China
• National grid projects are followed by industrial grid
platform development by IBM, Microsoft, Sun, HP,
Dell, Cisco, EMC
• Grid service providers (GSPs) and new grid
applications have emerged rapidly, similar to the
growth of Internet and web services
• Grid systems are classified in essentially two
categories:
computational or data grids and P2P grids

Peer-to-Peer Network Families

• A well-established distributed system is the client- Structure of P2P
server architecture
• P2P network is client-oriented instead of server-
oriented.
• Every node acts as both a client and a server.
• Peer machines are simply client computers
connected to the Internet.
• All client machines act autonomously to join or
leave the system freely.
• No master-slave relationship exists among the
peers.
• No central coordination or central database is
needed

4
 9/1/2024

• The peers are totally unrelated. • When a new peer joins the system, its peer ID is
• Each peer machine joins or leaves the P2P added as a node in the overlay network.
network voluntarily. • When an existing peer leaves the system, its peer
ID is removed from the overlay network
• Only the participating peers form the physical
automatically.
network at any time.
• There are two types of overlay networks:
• A P2P network does not use a dedicated unstructured and structured.
interconnection network
• An unstructured overlay network is characterized
• The physical network is formed using the TCP/IP by a random graph.
and NAI protocols • There is no fixed route to send messages or files
Overlay Networks among the nodes
• Based on communication or file-sharing needs, • Flooding is applied to send a query to all nodes in
the peer IDs form an overlay network at the an unstructured overlay
logical level

• 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
• P2P networks are classified into four groups
Issues in P2P
1.Fault tolerance, failure management, load balancing,
2.Lack of trust among peers
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

5
 9/1/2024

NIST Cloud Computing Reference Architecture

• National Institute of Standards and Technology
(NIST).
• The model offers approaches for secure cloud
adoption while contributing to cloud computing
guidelines and standards.
• The NIST team works with leading IT vendors,
developers of standards, industries and other
governmental agencies at a global level to
support effective cloud computing security
standards and their further development

• Cloud is a parallel and distributed computing system • Each actor is an organization plays an important role in a
consisting of a collection of inter- connected and transaction or a process, or performs some important
virtualized computers that are dynamically provisioned task in cloud computing
and presented as one or more unified computing • A cloud consumer can request cloud services directly
resources based on service-level agreements (SLA) from a CSP or from a cloud broker.
established through negotiation between the service Interactions between different actors in a cloud
provider and consumers.
• The cloud reference architecture includes five major
actors :
Cloud consumer
Cloud provider
Cloud auditor
Cloud broker
Cloud carrier

6
 9/1/2024

Cloud Consumer Usage scenarios

• The cloud consumer uses the services from a CSP or
person or asks an organization that maintains a
business relationship.
• The consumer then verifies the service catalogue
from the cloud provider and requests an appropriate
service or sets up service contracts for using the
service.
• The cloud consumer is billed for the service used.

Example of cloud services available to cloud consumers Cloud Provider

• Cloud provider offers cloud services to interested
parties.
• A cloud provider manages the infrastructure needed
for providing cloud services.
• The CSP also runs the software to provide services and
organizes the service delivery to cloud consumers
through networks.
• SaaS providers then deploy, configure, maintain and
update all operations of the software application on
the cloud infrastructure

7
 9/1/2024

• PaaS cloud providers manage the computing

Major activities of a cloud provider
infrastructure and ensure that the platform runs the
cloud software and implements databases,
appropriate runtime software execution stack and
other required middleware elements.
• They support development, deployment and the
management of PaaS consumers by providing tools
such as IDEs, SDKs and others.
• The IaaS CSP aggregates physical cloud resources such
as networks, servers, storage and network hosting
infrastructure

Cloud Auditor Cloud Broker

• It collects service requests from cloud consumers and
• The cloud auditor performs the task of independently manages the use, performance, and delivery of cloud
evaluating cloud service controls to provide an opinion services
when requested. • Also negotiate and manage the relationship between
cloud providers and consumers
• Cloud audits are done to validate standards A cloud broker may provide services that fall into one of the
conformance by reviewing the objective evidence. following categories
• The auditor will examine services provided by the cloud • Service intermediation : will improve some specific
capabilities, and provide value added services to cloud
provider for its security controls, privacy, performance consumers.
• Service aggregation : links and integrates different
services into one or more new services.
• Service Arbitrage : similar to aggregation, except for the
fact that services that are aggregated are not fixed.
In service arbitrage, the broker has the liberty to
choose services from different agencies.

8
 9/1/2024

Cloud Carrier Cloud Deployment Models

• Tries to establish connectivity and transports cloud The NIST have classified cloud deployment
services between a cloud consumer and a cloud models into four categories namely,
provider.
 Public cloud
• It offer network access for consumers, by providing
telecommunication links for accessing resources using  Private cloud
other devices  Hybrid cloud
• It provide secure and dedicated high - speed links with  Community cloud
cloud providers and between different cloud entities.

Public Cloud Advantages of public cloud

• The public cloud services are runs over the 1. It saves capital cost behind purchasing the server
internet. hardware’s, operating systems and application software
• The users who want cloud services need internet licenses.
connection in their local device like thin client, thick 2. There is no need of server administrators to take care of
client, mobile, laptop or desktop etc. servers as they are kept at CSPs data center and managed
by them.
• The public cloud services are managed and
maintained by the Cloud Service Providers (CSPs) or 3. No training is required to use or access the cloud
the Cloud Service Brokers (CSBs). services.
4. There is no upfront or setup cost is required.
• Pricing like subscription or pay per- use model
5. A user gets easy access to multiple services under a
• Users can easily access the services without single self - service portal.
purchasing any specialize hardware or software
6. Users have a choice to compare and select between the
providers.

9
 9/1/2024

Private Cloud
Disadvantages of public cloud
• Used by the organizations internally.
1. There is lack of data security as data is stored on
public data • It run over the intranet connection.
2. Expensive recovery of backup data. • Designed for a single organization, anyone within the
organization can get access to data, services and web
3. User never comes to know where (at which location) applications easily through local servers and local network
their data gets stored, how that can be recovered and
• Outside users of the organizations cannot access them.
how many replicas of data have been created.
• The infrastructure is fully managed and maintained by the
organization itself.
• It is much more secure than public cloud.
• It also provides good level of trust and privacy to the users.
• More expensive than public clouds due to the capital
expenditure involved in acquiring and maintaining them.

The well-known private cloud platforms are Openstack, Open Hybrid Cloud
nebula, Eucalyptus, VMware private cloud etc • Composed of two or more clouds that offers the
Advantages of private cloud benefits of multiple deployment models.
1. Speed of access is very high as services are provided • It mostly comprises on premise private cloud and off
through local servers over local network. premise public cloud to leverage benefits of both and
2. It is more secure than public cloud as security of cloud allow users inside and outside to have access to it.
services are handled by local administrator. • The Hybrid cloud provides flexibility such that users can
3. It can be customized as per organizations need. migrate their applications and services from private
4. It does not require internet connection for access. cloud to public cloud and vice versa
5. It is easy to manage than public cloud. The popular hybrid clouds are AWS with eucalyptus, AWS
Disadvantages of private cloud with VMware cloud, Google cloud with Nutanix etc.
1. Implementation cost is very high • The limitations of hybrid cloud are compatibility of
2. It requires administrators for managing and maintaining
deployment models, vendor-lock in solutions, requires a
servers. common cloud management software

10
 9/1/2024

Community Cloud Cloud Service Models

• Combination of one or more public, private or • A Cloud computing is meant to provide variety
hybrid clouds, which are shared by many of services and applications for users over the
organizations for a single cause. internet or intranet.
• The community cloud is setup between multiple • Three service classes which are called cloud
organizations whose objective is same. service models or cloud reference models or
• The Infrastructure for community cloud is to be working models of cloud computing
shared by several organizations within Infrastructure as a Service (IaaS)
The well-known community clouds are Salesforce, Platform as a Service (PaaS)
Google community cloud etc.
 Software as a Service (SaaS)

The cloud computing stack

11
 9/1/2024

Infrastructure as a Service (IaaS) • Instead of purchasing extra servers, softwares,

• IaaS can be defined as the use of servers, storage, datacenter space or network equipment, IaaS
computing power, network and virtualization to enables on-demand provisioning of
form utility like services for users. computational resources in the form of virtual
• It provides hardware resources virtualization machines in cloud data center
• CSP owns all equipment, such as servers, storage • Providers of IaaS are Amazon Web Services (AWS),
disks, and network infrastructure. Microsoft Azure, GoGrid, Joyent, Rackspace etc.
• Developers can create virtual private storage, • Cloud softwares through which IaaS can be setup
virtual private servers, and virtual private are Openstack, Apache Cloud Stack, Eucalyptus,
networks by using IaaS and VMware VSphere etc.
• It is a way of delivering cloud computing
infrastructure as an on-demand service

Components in IaaS service model

• Customer has controls over the OS, storage

and installed applications, but has limited
control over network components
• User cannot control the underlying cloud
infrastructure.
• Services offered by IaaS include web servers,
server hosting, computer hardware, OS,
virtual instances, load balancing, web servers
and bandwidth provisioning.

12
 9/1/2024

Platform as a Service (PaaS) Some key providers of PaaS clouds are Google App
• Allows the user to create web applications quickly and Engine, Microsoft Azure, NetSuite, Red hat Open
easily and without worrying about buying and shift etc
maintaining the software and infrastructure
• PaaS provides tools for development, deployment and
testing the softwares, middleware solutions,
databases, programming languages and APIs for
developers to develop custom applications; without
installing or configuring the development environment.
• It is built on the top of IaaS layer.
• The challenge is compatibility i.e. if user wants to
migrate the services from one provider to other then
they have checked the compatibility of execution
engine and cloud APIs first.

• Users interact with the software and append and Power of PaaS
retrieve data, perform an action, obtain results from • Source code development : provide the users with a
a process task and perform other actions allowed by wide range of language choices including stalwarts
such as Java, Perl, PHP, Python and Ruby.
the PaaS vendor
• Websites : provide environments for creating,
• Customer does not own any responsibility to running and debugging complete websites, including
maintain the hardware and software and the user interfaces, databases, privacy and security tools.
development environment • Developer sandboxes : PaaS also provides dedicated
“sandbox” areas for developers to check how
Example: snippets of a code perform prior to a more formal
Google app engine, Salesforce.com, Google Earth, test.
Google Maps, Gmail • Sandboxes help the developers to refine their code
quickly and provide an area where other
programmers can view a project, offer additional
ideas and suggest changes or fixes to bugs.

13
 9/1/2024

Complications with PaaS Compatibility : Most businesses have a restricted set of

Interoperability : PaaS works best on each provider’s own programming languages, architectural frameworks and
databases that they deploy.
cloud platform, But the risk here is that the customizations
• It is important to make sure that the vendor you
or applications developed in one vendor’s cloud may not be choose supports the same technologies.
compatible with another vendor and hence not necessarily
For example, if you are strongly dedicated to a .NET
migrate easily to it. architecture, then you must select a vendor with
• Customers agree with being hooked up to a single native .NET support.
vendor, this may not be the situation every time. Users • Likewise, database support is critical to performance
may want to keep their options open. and minimising complexity.
• In this situation, developers can opt for open - source Vulnerability and security : Multitenancy lets users to
solutions. Open - source PaaS provides elasticity by be spread over interconnected hosts.
revealing the underlying code and the ability to install the • The providers must take adequate security measures
in order to protect these vulnerable hosts from
PaaS solution on any infrastructure. attacks, so that an attacker is not able to easily access
• The disadvantage of using an open source version of PaaS the resources of host and also tenant objects.
is that certain benefits of an integrated platform are lost.

Vendor lock-in : Pertaining to the lack of Software as a Service

standardisation, vendor lock-in becomes a key • Software-as-a-Service is designed for on demand
barrier that stops users from migrating to cloud applications or software delivery to the cloud
services. users.
• Lock-in issues arise when a company decides to • User does not need to install required software in
change cloud providers but is unable to migrate
their local device as they are provided remotely
its applications or data to a different vendor.
through network.
• Most customers are unaware of the terms and
conditions of the providers that prevent • The consumer of a SaaS application requires thin
interoperability and portability of applications. client software such as a web browser to access
• A number of strategies are proposed on how to the cloud-hosted application.
avoid/lessen lock-in risks before adopting cloud Some examples of popular SaaS applications are
computing. Google Docs, Hotmail, Salesforce and Gmail

14
 9/1/2024

• SaaS applications come in a variety of applications to Architectural design of compute and storage clouds
include custom software such as CRM applications, Cloud Platform Design Goals
Helpdesk applications, HR applications, billing and
invoicing applications • Scalability, virtualization, efficiency, and reliability are
four major design goals of a cloud computing platform
• The APIs allow modifications to the security model,
data schema, workflow characteristics and other • Clouds support Web 2.0 applications
functionalities of services as experienced by the user • The cloud management software needs to support both
Examples of SaaS platform enabled by APIs include physical and virtual machines
Salesforce.com, Quicken.com • The platform needs to establish a very large-scale
SaaS provides the following services : HPC(High performance computing) infrastructure.
• Enterprise - level services • If one service takes a lot of processing power, storage
• Web 2.0 applications including social networking, capacity, or network traffic, add more servers and
blogs, wiki servers, portal services, metadata bandwidth.
management and so on. • Data can be put into multiple locations.

Enabling Technologies for Clouds

• Broadband and wireless networking, falling storage costs,
and progressive improvements in Internet computing
software.
• Users are able to demand more capacity at peak demand,
reduce costs, experiment with new services, and remove
unneeded capacity,
• Service providers can increase system utilization via
multiplexing, virtualization, and dynamic resource
provisioning.

15
 9/1/2024

A Generic Cloud Architecture

• The Internet cloud is envisioned as a massive cluster
of servers.
• These servers are provisioned on demand to perform
collective web services or distributed applications
using data-center resources.
• The cloud platform is formed dynamically by
provisioning or deprovisioning servers, software, and
database resources.
• Servers in the cloud can be physical machines or VMs.
• Other cloud resources are added into a cloud
platform, including storage area networks (SANs),
database systems, firewalls, and security devices

• The software infrastructure of a cloud platform must

handle all resource management and do most of the • Private clouds are easier to manage, and
maintenance automatically. public clouds are easier to access.
• Software must detect the status of each node server • The trends in cloud development are that
joining and leaving, and perform relevant tasks
accordingly. more and more clouds will be hybrid.
• Cloud computing providers, such as Google and • Security becomes a critical issue in
Microsoft, have built a large number of data centers safeguarding the operation of all cloud types
all over the world.
• Each data center may have thousands of servers.
• The location of the data center is chosen to reduce
power and cooling costs.
• Thus, the data centers are often built around
hydroelectric power.

16
 9/1/2024

Layered Cloud Architecture

• A cloud is developed at three layers:

infrastructure, platform, and application
• Infrastructure layer is deployed first to support
IaaS services.
• The infrastructure layer is built with virtualized
compute, storage, and network resources

• The platform layer is for general-purpose and repeated Market-Oriented Cloud Architecture
usage of the collection of software resources.
• Market-oriented resource management is necessary
• This layer provides users with an environment to to regulate the supply and demand of cloud
develop their applications, to test operation flows, and resources to achieve market equilibrium between
to monitor execution results and performance supply and demand.
• The application layer is formed with a collection of all • The designer needs to provide feedback on economic
needed software modules for SaaS applications. incentives for both consumers and providers, to
• Applications in this layer include daily office promote QoS-based resource allocation mechanisms
management work, such as information retrieval, • Users or brokers acting on user’s behalf submit
document processing, and calendar and authentication service requests from anywhere in the world to the
services. data center and cloud to be processed
• The application layer is also used by enterprises in • The SLA resource allocator acts as the interface
business marketing and sales, consumer relationship between the data center/cloud service provider
management (CRM), financial transactions, and supply and external users/brokers
chain management

17
 9/1/2024

• The VM Monitor mechanism keeps track of the

availability of VMs and their resource entitlements.
• The Dispatcher mechanism starts the execution of
accepted service requests on allocated VMs.
• The Service Request Monitor mechanism keeps track
of the execution progress of service requests.

Design Challenges Challenges related to Unpredictable Performance

Challenges related to Data Privacy, Compliance and and Bottlenecks
Security Concerns
• Multiple Virtual Machines (VMs) shares the
• The most common attacks on the network include resources with each other like CPU, memory, I/O
buffer overflows, DoS attacks, spyware, malware, root
kits, trojan horses and worms.
and network
• By encrypted data, virtual LANs and network • It may generate an unpredicted performance and
middleboxes such as firewalls, packet filters etc., many may result into system bottlenecks
challenges can be solved immediately. • One solution for this challenge is to improve I/O
• Newer attacks may result from hypervisor malware, architectures and operating systems used in
guest hopping and hijacking or VM rootkits in a cloud physical servers
environment
• Encrypt their data before placing it in a cloud.

18
 9/1/2024

Challenges related to Cloud Scalability, Interoperability and

Challenges related to Service Availability and Standardization
Vendor/Data Lock-in
• In cloud computing, pay-as-you-go model refers to utility -
• Service availability may result into huge financial loss. based model where bill for storage and the bandwidth of the
• In such case, even if a company has multiple data network are calculated according to the number of bytes
centers located in different geographic regions, it used.
may have common software infrastructure and • Google App Engine scales and decreases automatically in
accounting systems. response to load increases; users are paid according to the
cycles used.
• Use of multiple cloud providers may provide more
protection from failures. • Amazon Web Service charges the number of instances used
for VM by the hour, even though the computer is idle.
• Software stacks can be used to enhance
• The cloud platform should also introduce live cross -
interoperability between various cloud platforms as platform migration between x86 Intel and AMD technologies
well as standardize APIs to rescue data loss due to a and support legacy load balancing hardware to avoid the
single company failure challenges related to interoperability

Challenges related to Software Licensing and Reputation Challenges related to Distributed Storage and Bugs in
Sharing Software
• The cloud computing providers primarily depended • In cloud applications the database services continuously
on open source software, as the commercial grow, Need efficient distributed SANs.
software licensing model is not suitable for utility • A major problem in cloud computing is data consistency
computing testing in SAN - connected data centers.
• Bad conduct by one client can affect the credibility of • Large - scale distributed bugs cannot be replicated, so
the cloud as a whole. debugging must take in the data centers for production.
For example, In AWS, spam - prevention services can • Debugging on simulators is another way to fix the
restrict smooth VM installation by blacklisting of EC2 problem, if the simulator is well designed.
IP addresses.

19