Distributed
Computing
11/20/2023 1
Distributed Computing
• A distributed computer system consists of multiple software components that are
on multiple computers, but run as a single system.
• The computers that are in a distributed system can be physically close together and
connected by a local network, or they can be geographically distant and connected
by a wide area network.
• A distributed system can consist of any number of possible configurations, such as
mainframes, personal computers, workstations, minicomputers, and so on. The goal
of distributed computing is to make such a network work as a single computer.
11/20/2023 2
Distributed Computing
• There are several advantages of this system over traditional
computing models such as scalability, reliability, flexibility.
• Distributed computing has many forms such as cluster
computing, grid computing ,cloud computing etc.
11/20/2023 3
Cluster Computing
• Cluster computing involves connecting multiple computers
(nodes) together to work as a single integrated computing
resource. These nodes are usually connected in a local area
network (LAN) and are used to solve complex computational
tasks.
11/20/2023 4
Cluster Computing
Characteristics of this computing:
❑ Many similar machines, close interconnection
❑ Often special, standardized hardware (racks, blades)
❑ Usually owned and used by a single organization
However, these cluster of machines require a massive amount of power and
cooling mechanism. Therefore, these clusters are moved into separate
building known as data center
11/20/2023 5
Grid Computing
• Grid computing is a form of distributed computing where resources from
various sources are combined to solve big computational problems.
• It allows resources to be shared and combined across computers located in
diverse geographical areas and networks.
11/20/2023 6
Grid Computing
• Grids use special software (middleware) to manage and distribute
tasks across diverse and widely spread resources and authorize
any process that is being executed on the network.
• Grids are typically used for scientific, engineering, and research
applications that require substantial computational power.
11/20/2023 7
Cloud Computing
• Cloud computing is a model for enabling convenient,
on-demand network access to a shared pool of configurable
computing resources that can be rapidly provisioned
and released with minimal management effort or
service provider interaction.
• It provides a wide range of services that can be distinguished into
3 catogery: SaaS, PaaS and IaaS
11/20/2023 8
What is virtualization?
Bob
Charlie
Alice
Physical machine
Daniel
■ Suppose Alice has a machine with 4 CPUs and 8 GB of
memory, and three customers:
■ Bob wants a machine with 1 CPU and 3GB of memory
■ Charlie wants 2 CPUs and 1GB of memory
■ Daniel wants 1 CPU and 4GB of memory
■ What should Alice do?
11/20/2023 9
What is virtualization?
Bob
Virtual
machine
monitor Charlie
Alice
Physical machine
Virtual machines Daniel
■ Alice can sell each customer a virtual machine (VM)
with the requested resources
■ From each customer's perspective, it appears as if they had a
physical machine all by themselves (isolation)
11/20/2023 10
What is virtualization?
• Virtualization is a technique to separate a service from the
underlying physical delivery of that service.
• Virtualization allows sharing of a single physical instance of a
resource or an application among multiple customers and
organizations at one time.
• It is a key characteristics of cloud computing that provides the
required isolation, abstraction and encapsulation
11/20/2023 11
Security Issues
• Cloud computing provides improved, optimized, and low cost
services to the customers through virtualization and multi-
tenancy architecture.
• The use of cloud technologies, service models, and deployment
methods brings about specific security risks and vulnerabilities
unique to the cloud environment.
11/20/2023 12
Security Issues
• Issues at communication level :
We divide the cloud communication into two categories, namely:
(a) communication external to the cloud
(b) communication internal to the cloud
The challenges faced by the external communication are same as
the challenges of conventional IT communication.
11/20/2023 13
Security Issues
• Shared communication infrastructure:
oThe sharing of network components provides attacker the window
of cross-tenant attack.
oService providers often restrict users from conducting
vulnerability scans to avoid confusion with potential cyberattacks,
ensuring the security and stability of their networks.
11/20/2023 14
Security Issues
• Issues at Virtual network:
o The virtual networks are responsible for communication between VMs.
Security and protection mechanisms over the physical network are not
able to monitor the traffic over virtualized network.
o Intrusion detection and prevention mechanisms usually depend on the
traffic patterns and activities to judge the anomalies and detect the
possibility of the attack. Virtualized network poses a hindrance to the
goal of such preventive measures
11/20/2023 15
Security Issues
• Security misconfigurations:
oA small misconfiguration can breach the security of the system.
oOne of the most common misconfiguration occur when
administrators select such a configuration tool that they are
familiar with but not necessarily covers all the security
requirements.
oThis can lead to data leakage of private information of customers.
11/20/2023 16
Security Issues
As a security measure for these attacks, Advanced cloud protection
system has been proposed where :
• VMs are constantly monitored.
• auditability for the actions of VMs.
• interceptor module is used for detecting any suspicious activities
To avoid misconfigurations, workforce needs to be extensively trained
with appropriate knowledge of the entire framework.
11/20/2023 17
Security Issues
• Virtualization issues:
o Virtualization can lead to data breach and cross-VM attacks.
Isolation is required storage devices, memory and hardware level.
oDuring the migration phase, the contents of VM as well as the
source code of the VM itself becomes exposed to the network.
oA successful VM escape attack can provide access to the
computing and storage hardware
11/20/2023 18
Security Issues
• Virtualization issues:
o The key module of virtualization is hypervisor or VMM. A
compromised VMM can put all the VMs that are managed by the
victim VMM under attacker’s control. The metadata of the VMs,
kept by the VMM, may also be exposed to an attacker if the
attacker takes control of a VMM.
11/20/2023 19
Security Issues
• Strategies to mitigating thse risks:
o The stored VMs should be encrypted.
oBuilt in security measures should be adopted for virtualized OS
oHypervisor state monitoring through third party
oThird party security technology should be used
11/20/2023 20
Security Issues
• Data/storage issues:
oData privacy and integrity
oData recovery vulnerability
oData backup
The users and organizations are encouraged to use standard
encryption algorithms with off-the-shelf-technology.
11/20/2023 21
Reliability in CC
• Reliability in cloud computing is how consistently a cloud computing system
is able to provide its services without interruption and failure.
• Cloud computing is a service-oriented architecture so the attributes of the
reliability rely on service models
• If a failure event has occurred at a particular instance, then partial or even
delayed services need to be delivered. The recovery should be done with
minimum intervention of human.
11/20/2023 22
Reliability in CC
Service Failure Management in Cloud Computing:
Ensuring reliable services in cloud computing involves effectively handling
service failures. Service failure management techniques are categorized into
two groups:
1. Reactive Failure Management and
2. Proactive Failure Management
11/20/2023 23
Reliability in CC
Reactive Failure Management : In this model, measures are taken after
the occurrence of failure. Whenever failures have occurred in cloud services,
the required measures will be taken by restarting the services from the last
execution instance recorded earlier using checkpointing or logging.
11/20/2023 24
Reliability in CC
Checkpointing:
• In this method, current state of a running process is saved on some backup
resources and on the occurrence of failure, the process will be restarted or
rolled back by using the last saved state.
• It has been proved that using checkpointing task competition time can be
reduced to linear from exponential.
11/20/2023 25
Reliability in CC
Checkpointing:
• On the basis of the working principle, checkpointing has divided been into
three different categories such as Uncoordinated Checkpointing, Coordinated
Checkpointing and Communication Induced Checkpointing.
• The determination of optimal checkpointing interval is a crucial design
decision for system performance.
11/20/2023 26
Reliability in CC
Logging:
• Each process is recorded or saved in its present state and messages are sent
periodically as the logs at some stable storage. When a process crashes, a new
process is created on the place of a crashed process by using the recorded
logs.
• The logs are saved in the order of the operations execution for regeneration
of a process in case of a failure.
11/20/2023 27
Reliability in CC
Logging:
• In the event of a failure, the system can use the log to either replay the
operations that were lost or undo the operations that were only partially
completed. This ensures that the system can recover to a consistent state and
avoid orphan process.
• To avoid logging overheads, system can combine checkpointing and logging.
11/20/2023 28
Reliability in CC
Redundancy:
Here, the backup resources are used to run replicas of the running processes.
On the basis of updating of running replicas to handle the inconsistency,
replication has divided into two categories :
• Primary Backup (Passive) replication
• Active replication
11/20/2023 29
Reliability in CC
Proactive Failure Management :
• In this method, the prevention measures are taken before the occurrence of
failure.
• The productivity of proactive failure management methods depends upon the
prediction of the occurrence of the failures.
• On the basis of the failure prediction results, the running processes are
migrated from the suspected resource to other healthy resource for an
uninterruptable execution.
11/20/2023 30
Reliability in CC
Proactive Failure Management :
• Migration method is used to provide fault tolerance by incorporating failure
prediction methods.
• Due to the dynamic nature of the cloud infrastructure, VM migration based
fault-tolerance methods more popular.
• There are 2 approaches for VM migration: Pre-copy VM Migration Approach
and Post-copy VM Migration Approach.
11/20/2023 31
Reliability in CC
Pre-copy VM Migration Approach :
• Hypervisor copies the state of the running VMs such as CPU state, memory
state, and state of other devices and the VM stops at the source machine. This
is the warm up phase.
• The stop and copy phase copies the remaining les or pages (if any) in the
memory that gets modfied (dirty pages) during the warm-up phase.
• After the transfer of all the pages the virtual machine resumes its execution
over the destination machine
11/20/2023 32
Reliability in CC
Post-copy VM Migration Approach :
• The process involves suspending running virtual machines (VMs) at the source nodes and then
migrating them to the destination nodes.
• During this migration, only partial attributes of the execution state are transferred, including
details such as CPU state and register usage. At the destination, the VMs resumes with the
execution.
• In parallel the source machine also stay active serving the migrated VMs.
• Whenever a VM does not find a page in its local memory, destination machine redirects the
page request to the source machine which in-turn responds with the faulted page
11/20/2023 33
Green ICT
Green IT (green information technology) is the practice of creating and using
environmentally sustainable computing resources.
Green IT aims to minimize the negative effects of IT operations on the environment by
designing, manufacturing, operating and disposing of servers, PCs and other computer-
related products in an environmentally friendly manner.
The motives behind green IT practices include reducing the use of hazardous materials,
maximizing energy efficiency during a product's lifetime, and promoting the
biodegradability of unused and outdated products.
11/20/2023 34
Green ICT
Importance of green ICT:
• Climate change
• Compliance
• Competitive advantage
• Reduced use of energy and other natural resources
11/20/2023 35
Green ICT
Challenges of green ICT:
1. Cost
2. Cultural pushback
3. Prioritization
4. Conflicting initiatives
5. Emerging fields
11/20/2023 36
Reference
• Web
• Security in cloud computing: Opportunities and challenges Mazhar Ali, Samee U. Khan,
Athanasios V. Vasilakos
• Reliability and Energy Efficiency in Cloud Computing Systems: Survey and Taxonomy-
Yogesh Sharma, Bahman Javadi, Weisheng Si,Daniel Sun
• https://www.techtarget.com/searchcio/definition/green-IT-green-information-
technology
11/20/2023 37
