2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

A Survey of Virtual machine migration, Optimal

Resource Management and Challenges
Kadu N.B. Dr.Pramod Jadhav Nirmal M.D.
2022 5th International Conference on Contemporary Computing and Informatics (IC3I) | 979-8-3503-9826-7/22/$31.00 ©2022 IEEE | DOI: 10.1109/IC3I56241.2022.10072618

Dr. A. P. J. Abdul Kalam University Associate Professor Dr. A. P. J. Abdul Kalam University
Indore, Madhya Pradesh, India Dr. A. P. J. Abdul Kalam University Indore, Madhya Pradesh, India.
[email protected] Indore, Madhya Pradesh, India. [email protected]
[email protected]

Abstract - Virtual machines (VMs) have become

increasingly valuable for resource consolidation and
management due to their efficient and secure containers, as
well as their capability to offer desired execution environments
for applications. The cloud is also becoming increasingly
popular, which is hosted in large data centers. The vast
majority of these large data centers use a virtualized server
infrastructure, known as a virtual machine (VM), that is
managed by a cloud infrastructure service such as an open
stack or cloud stack, etc. Computing, storage, and
communication resources are increasingly needed in large
Cloud Data Centers (CDCs). With the growth of cloud
environments, VM migration has become an important
criterion to ensure Quality of Service (QoS), save energy, and
reduce resource usage. This paper examines the merits,
challenges, and resource management aspects of virtual Figure.1: VM Migration in Cloud Computing
machines migration. In addition to providing load balancing,
online system maintenance, proactive fault tolerance, power Virtualization allows each execution environment,
management, and resource sharing, it also helps to manage which is known as a Virtual Machine (VM), to function
system power consumption. independently while simultaneously running an OS and an
application without interfering with one another. This is a
Keywords: Virtual machine migration, Resource critical aspect of cloud computing (OS). As shown in Figure
Management, Cloud Data Center, Cloud computing, 2, virtualization divides hardware resources into numerous
Virtualization, Live migration
distinct execution contexts. By using multiple instances of
the host, users can manage multiple virtual machines on the
I. INTRODUCTION
host. As a result of virtualization, virtual computers can be
moved from one physical host to another, enhancing
Data innovation with cloud computing involves
efficiency (e.g., lower costs) and enabling flexibility (such
accessing resources from the Internet via online tools and
as the ability to share CPU, bandwidth, memory). In the case
programs, as opposed to keeping data on a server. Instead of
of resource shortages, a virtual machine hosted on the server
storing documents locally on a hard drive, you can use
is moved to another server with sufficient resources.
cloud- based storage to save them to a remote database.
Using cloud computing technology, reps are able to access
and retrieve data from anywhere and from any device. It is
also known as "advanced computing" and "big data.". The
cloud allows users to store and access data remotely, and the
data can be accessed through the Internet. Clouds may be
classified as public, private, hybrid, or community clouds,
depending on their location.

The terms Infrastructure-as-a-Service (IaaS) and

Platform-as-a-Service (PaaS) refer to the delivery of
computing infrastructure by the cloud provider to the client
in the form of virtual machines (VMs), respectively; SaaS
(Software-as-a-Service) refers to the delivery of software
applications by the cloud provider. Based on Figure 1, all
Figure.2: Virtual Machine
services are provided to consumers using a pay-as-you-go
model, no matter where they are situated. There has been an increase in the capabilities of IT
resources over the past few years, which is attracting app
developers and service providers to use them. Researchers

450
979-8-3503-9826-7/22/$31.00 ©2022 IEEE

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

and service providers distribute these resources among end place during the initial phase of virtual machine migrations.
users to optimize their use and to maximize their profits. Cloud computing task-based scheduling algorithms have
Furthermore, these resources have high processing speeds now incorporated particle swarm optimization methods for
and large storage capacities. Various cloud computing task- based system load balancing, as proposed by [2014]
environments are available including Google App Engine [6]. It transferred excess workload from overburdened
[2], Microsoft Azure [3], Amazon EC2 [4], IBM Smart virtual machines instead of transferring the entire machine.
Cloud [5], etc. In cloud computing, these environments are A proposed agent-based method for commodity load
provided in the form of hardware, software, storage, balancing was published in 2015[7], titled "Agent-based
platform, infrastructure, databases, and much more. On- load balancing in Cloud data centers".
demand resources can be supplied by cloud computing by
using a pay- per-use [6] model. This is accomplished The queuing theory was used to illustrate how virtual
through the delivery of hardware and software services over machines interacted, thus decreasing migration. The
the internet. Queuing Theory Based Bandwidth Allocation Algorithm for
Live Virtual Machine Migration, 2015[8]. Performance of a
Cloud service providers now manage warehouse-sized virtual computer was improved as a result of the approach.
data centers due to the increased demand for cloud [2009]
resources. A large amount of electricity is used by the Cloud [6] suggests a strategy for optimizing energy consumption in
Data Center (CDC), which is home to thousands of cloud computing environments by moving workloads and
computing servers connected by fast connections. Servers distributing tasks among virtual machines. As part of the
typically have idle times of around 30%, and only 10% to workload, virtual machines must be moved. A Bee colony
15% of their total capacity is actually used to satisfy approach for task load balancing in a cloud computing
resource demands, making it impossible to deliver services environment is described in [2015] [7], in which virtual
at the levels promised. Increases in operational expenses and computers in queues are assigned tasks that have the highest
energy use are astounding when resources are under- or priority. The algorithm improves the customer experience
over-provisioned. According to estimates, CDC's operating (QOS).
expenditures would account for 75% of infrastructure and
power use in 2014, while IT would only account for 25% It was shown that it is necessary to use a distinct
[10, 11]. More than 200,000 houses could have been statistical distribution to represent CPU, memory, and total
continually powered by the 260 million Watts of electricity amount of instructions to facilitate the simulation of
used by Google data centers in 2013.Switching idle mode resource usage in cloud computing [2015] [10]. The model
servers to sleep or off mode depending on resource demands explained the necessity to employ a distinct statistical
is one of the basic fixes for these problems. Energy is distribution to represent CPU, memory, and total amount of
effectively conserved in this way. instructions in cloud computing. The hybrid meta-heuristic
method was proposed to schedule VMs with load balancing
II. RELATED WORK in cloud computing. [2014] [11] A hybrid meta-heuristic
approach incorporated both particle swarm optimization and
Based on anticipated future requests from virtual ant colony optimization. To speed up virtual machine
machines, an evaluation of resource allocation status is migration and reduce computation time, the PSO operator
updated; however, the results are unreliable in anticipating was added to the ACO algorithm. Data center migration
dynamic demands and practical needs. It evaluates the using network-aware load balancing [2014][12]. An
power topology of the data center and takes into account approach to multi-resource load balancing that is network-
intricate power models in order to implement this strategy. aware and achieves consistent performance.
[11]. By [2018] [3], a dynamic virtual machine
consolidation technique for cloud data centers was created. A bin packing technique that dynamically links and
The deployment and migration of virtual machines were disconnects virtual machines based on the speed of the
proposed to use Space Aware Best Fit Decreasing (SABFD). displayed resource has been successfully used to perform
Energy-aware dynamic VM consolidation for green cloud dynamic load balancing in the cloud. [2016] [13]. Migration
computing will become even more alluring with the option of cloud data centers using network-aware, self-adaptive
of VM selection. The process of allocating resources with a virtual machines [2016]. Artificial intelligence with a
low energy consumption will become more attractive in reinforcement learning foundation for network-aware virtual
cloud computing environments [2009] [2]. It was suggested machines. This is partially due to the tendency of cloud data
that an estimation module be incorporated to estimate the centers to overlook energy efficiency and VM allocation.
system's impending load, followed by two schedulers to For multidimensional variable-sized virtual machine
handle anticipated and unexpected loads. Column generation allocation and migration at cloud data centers, Exact [2017]
is used by the scheduler when optimizing integer [17] has developed a revolutionary method. A large cloud
linear/quadratic programming. data center cannot use it because its processing is time-
Cloud computing load balancing and scheduling consuming. [2010][16] describes a scheduling strategy
technique utilizing live migration. Ant colonies will be based on evolutionary algorithms for load-balancing virtual
optimized using live migration. State monitoring will take machines

451
Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

in a cloud computing environment. This method, which 3) Workload characteristics: The effect that active
comes after schedule optimization, eliminates load processes in the migrated VM are having on migration
inequality and excessive migration expenses. speed. The programmes may, for instance, be heavily CPU-,
network-, or i/o-intensive. They will fight each other for the
III. OPTIMAL RESOURCE MANAGEMENT relevant resources using the migration strategy.
As shown in Figure 3, virtual resource management
collects performance data from physical servers, virtual 4) Host performance: The migration procedures
machines running different operating systems, and frequently rely on the resources of the source and
virtualized software and manages it centrally. By comparing destination servers, thus these servers are also crucial for
performance data for virtual machines with data for the real migration performance.
server, judgments can be made about the server's resources
and help optimize it. 5) Co-placed VMs: This option specifies the total
number of virtual machines that are co-located on the source
• It shows CPU, memory, and disc consumption and destination servers, as well as the total amount of
information for the physical server. This allows hardware that these machines used. Performance now only
monitoring of CPU, memory, and disc benefits from resource isolation, not virtualization. Running
consumption. on the same server, the performance of the other VMs would
suffer (along with the migration manner).
• For each guest's reports, the performance data of
the virtual machine is layered. This enables each 6) Migration example: Distinctive migration
guest's consumption of the CPU, RAM, and storage behaviors will result from various migration ideologies. A
to be viewed. notable quantity of advantages and overheads are also
introduced by special optimization approaches, similar to
this.

7) VM configuration: This refers to a virtual

machine's static parameters, such as the amount of memory
that is available, the amount of memory that is being used at
any one time, the size of the network buffer, the format of
the virtual disc, and other factors.

8) Disk i/o pattern: The pattern of the disc I/O similar

to memory access samples would impact performance, it
will slow down the migration of garbage data because dirty
disc blocks will need to be retransmitted.

Figure.3: Resource Management using Virtual Machine

9) Migration collection: When a large number of
virtual machines (VMs) are being migrated, either
The performance of VM migration will be impacted by sequential or parallel migration will produce a distinctive
a number of variables. The research that examines the performance.
relationships between migratory performances and related
issues is reviewed in this section. It can be listed in the 10) Virtual Machine Monitor (VMM): Although pre-
following order: reproduction migration is often utilized in VMMs, the
implementations in different VMMs are slightly different,
1) Memory enters the pattern: This shows the along with distinct migration results.
migrated virtual machine's memory access pattern and
memory dirtying rate. The performance of memory statistics 11) To reduce energy expenditures, resource allocation
migration is most significantly impacted by this element. is used in cloud computing: The research on reducing
The memory dirtying charge in particular affects whether energy expenses can be divided into two groups as well:
the pre-fresh reproduction's release section can find the right
termination factor. A. Bin packing to allocate resources statically

2) Migration bandwidth: This is the amount of B. Using thresholds to consolidate resources dynamically
bandwidth used to move data between servers at the source
and the destination. When compared to WAN environments, 1) Bin packing to allocate resources statically: Recent
the migration bandwidth in LAN contexts is much higher. studies show that a significant amount of the power required
by cloud computing systems is consumed by physical
machines (PM). Efficiency resource allocation, also known
as static resource allocation, is the process of allocating
VMs

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)
452

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

to PMs with the objective of minimizing the number of

active PMs. To express the static energy-aware resource
allocation problem in an understandable manner, VMs and
PMs can be modeled as the items and bins in the bin
packing problem, respectively. To create a polynomial
resource allocation technique for this transformed bin
packing problem with linear consumption costs, the lower
bound of the optimal solution (LB-OPT) was first
determined. The space division model for the
multidimensional bin packing problem was described while
accounting for the various PM resource types. Based on this
idea, a practical approach for allocating virtual machines
(VMs) was presented in order to lessen the uneven
consumption of the different resources.
Figure.4: Challenges in Virtual Machine
2) Using thresholds to consolidate resources
dynamically: Live migration technology, which enables the However, there are several issues with virtualization
transfer of VMs between PMs, has provided a solution for management that enterprises must deal with (Fig. 4). Here
dynamic resource consolidation. In order to allow the source are some typical virtualization difficulties and suggestions
PM to shut down without using any power during VM on how businesses might solve them.
migration, the system should, on the one hand, relocate VMs
from the low resource utilized source PM to another target Performance Tracking: It is necessary to monitor
PM. The system must, however, be cautious to avoid using virtual server performance in a different way than physical
the target PM excessively. The threshold-based resource server performance. Memory and CPU consumption cannot
consolidation technique has been looked into as a potential be measured using conventional techniques. Hardware
means of achieving these objectives. The high threshold (th) resources include the CPU, memory, and storage that are
and low threshold are the first two thresholds in this shared by all of the VMs in a virtual infrastructure. Each
technique that are predetermined (tl). The system will move VM shares a percentage of the resources available when a
some of the virtual machines (VMs) on a PM to another PM host supports hundreds of virtual machines, enabling
when that PM utilizes more resources than anticipated in concurrent data processing. To ensure that a virtual
order to avoid hotspots. The system will move all of the infrastructure is working, a specific set of performance
virtual machines that are now running on target PM pi to metrics must be examined. Metrics like CPU ready, memory
another target PM in order to save energy when pi's resource ready, memory bubble, and swapped memory require real-
usage falls below threshold (tl). All of these studies make time monitoring of all the VMs. Live VM transfer further
the assumption that there is a central manager who keeps complicates monitoring.
track of and retains data on each PM and VM in order to
accurately anticipate the two thresholds th and tl in dynamic Security: In virtualized systems, the hypervisor is
cloud systems. Allowing the PMs to manage resources in a the only security point of failure, and if it breaks, sensitive
distributed way as an alternative strategy can increase data is no longer protected. The amount of risk and exposure
system resiliency and avoid the need for a central manager. is greatly increased by this. For instance, "hyper-jacking" is
Recent advances in probability-based distributed VM the process of installing a malicious hypervisor that has the
consolidation methods have ignored live migration's capacity to acquire total control of a server. Because the
detrimental impact on QoS in favor of focusing primarily on operating system is blind to the infected machine, standard
the benefit of energy cost reduction. security procedures are worthless. Instead of corrective
action, preventative steps like hardening the environment
IV. RESEARCH CHALLENGES IN VIRTUAL MACHINE must be adopted. Therefore, hardware-rooted trust and safe
MIGRATION hypervisor launch are the best techniques for combating
hyper jacking. The risk of VM jumping, also known as
Virtualization, the use of software to mimic hardware guest- hopping, is more probable and just as harmful. The
capabilities for servers, networks, and data storage, has vast majority of the time, this attack technique takes use of
become a common practice for organizations in today's hypervisor flaws that let malware or remote attacks bypass
corporate culture. Virtualization has a lot of benefits over VM separation defenses and access other VMs, hosts, or
conventional hardware-based systems, such as lower costs, even the hypervisor itself. Attackers frequently carry out
quick and simple scaling, frequently less downtime, and these actions on the host after gaining access to a low-value,
increased functionality. hence less secure, VM. They then use that access as a
springboard for additional attacks on the system. It has
occasionally been

453

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

possible to attack secured VMs or the hypervisor itself by infrastructure must be constantly monitored. For the
working together with two or more compromised VMs. resource utilization of the virtual machines, a trend must be
found and projected over time. The server administrator can
Controlling VM Sprawl: Virtualization led to both utilize this data to understand the current pattern and
the uncontrolled proliferation of VMs and the excessive estimate how many resources will be required to handle the
resource allocation because of the speed with which storage expected rise in demand.
space could be quickly assigned to virtual machines (VMs)
(Fig.5). By not making the most of all the VMs that are
available, a conventional virtual infrastructure wastes
resources. As a result, VM sprawl has come to be seen as
virtualization's biggest flaw. Regular virtual infrastructure
monitoring is important to identify inactive VMs, ones with
over- or under-provisioned resources, and ones that haven't
even been powered on since they were established. This will
help minimize VM sprawl. By getting rid of these
extraneous VMs or optimizing the resources assigned to
them, virtual sprawl can be considerably minimized.
Ineffective hardware acquisitions are also avoided via
resource management.
Figure.6: VM Capacity planning

Licensing Issues: Software vendors provide a number of

different licensing options for virtualization. It's unfortunate
that there isn't a recognized approach for applying
measurements to virtual environments. Some software firms
build conversion models based on peak resource usage or
active instances, while others try to disregard the problem
and stay in the physical world. For example, one business
used a server with four physical cores and seven virtual
instances running on it. Depending on the vendor and the
programme they were using, they needed seven licenses
from one but only four from another.

Figure.5: VM Sprawl Challenges of Network based Live VM Migration:

Attacks on Virtualization Features: Virtual 1. Low Bandwidth across WAN: Because of the size
networking and VM migration are the most frequently of the images, transferring a live virtual computer
targeted characteristics of virtualization, despite the fact that across a low bandwidth, high latency WAN is
there are many other functions that might be misused. A impractical.
given VM could become completely susceptible to active
manipulation attacks and passive sniffing attacks as a result 2. Network error: In post copy, a VM will redirect to
of insecure VM migration. the source host while attempting to fetch pages that
haven't yet been transferred at the destination host,
Challenges with Compliance and System resulting in a network error since the VM is stalled
Management: When working with virtualized systems, (Fig.7).
compliance auditing and enforcement, as well as routine
system management, offer complex issues. Because of VM 3. Programs that make use of memory are called
sprawl and even inactive VMs, it will be challenging to memory-intensive programmes. This technique is
receive accurate results from vulnerability assessments, used to lessen the post copy migration approach's
patching/updates, and audits. overhead, and it performs better than reusing
memory concepts.
Capacity Planning: Virtual infrastructure, also
known as capacity planning, serves as the foundation for the 4. Live migration is used to control the transfer of
data centers that power today's businesses. As a business CPU and memory state between virtual distributed
grows, it must develop the infrastructure that underpins it. shared memory (VDSM) systems, which are
Without proper planning, the server management team groups of host machines.
would have trouble scaling the infrastructure to achieve
business goals (Fig.6). As part of capacity planning,
the virtual

454

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

migration cost model is now being investigated on a variety

of hardware platforms and virtualization technologies.

REFERENCES

[1] VMware Inc., vSphere resource management guide,2009

[2] How VMware virtualization right-sizes IT infrastructure to reduce
power consumption,. 2009
[3] Energy-Aware Dynamic Virtual Machine Consolidation for Cloud
Datacenters, IEEE transactions, volume 6, 2018.
[4] Wen, Wei-Tao, Chang-Dong Wang, De-Shen Wu, and Ying-Yan Xie,
“An ACO-Based Scheduling Strategy on Load Balancing in Cloud
Computing Environment,” In 2015 Ninth International Conference on
Frontier of Computer Science and Technology, pp. 364-369. IEEE,
2015.
Figure.7: Network Issues in VM [5] Ramezani, Fahimeh, Jie Lu, and FarookhKhadeerHussain, “Task-
based system load balancing in cloud computing using particle swarm
Post copy live virtual machine migration technique's optimization,” International Journal of Parallel Programming 42, no. 5
current challenges: pp. 739-754,2014
[6] Gutierrez-Garcia, J. Octavio, and Adrian RamirezNafarrate, “Agent-
based load balancing in Cloud data centers,” Cluster Computing 18,
1. The Post copy migration technique must have less no. 3 pp. 1041-1062, 2015.
overhead because it results in irregular page fault [7] Zhu, Changpeng, Bo Han, Yinliang Zhao, and Bin Liu, “A Queueing-
Theory-Based Bandwidth Allocation Algorithm for Live Virtual
generation when the destination host's virtual
Machine Migration,” In2015 IEEE International Conference on Smart
machine is running. Each time, the destination City/SocialCom/SustainCom (SmartCity), pp. 1065-1072. IEEE,
host's VM is placed in suspension while it waits for 2015.
the source host to deliver the required memory [8] Bo Li, Jianxin Li, JinpengHuai, TianyuWo, Qin Li,LiangZhong (2009),
“EnaCloud:AnEnegy-saving Application Live Placement Approach
pages once more. for Cloud Computing Enviorments” , In IEEE International Conference
on cloud Computing 2009, pp. 17-24.
2. It is necessary to reduce the amount of data transfer [9] Babu, KR Remesh, Amaya Anna Joy, and Philip Samuel, “Load
(memory pages and CPU state) between clusters of Balancing of Tasks in Cloud Computing Environment Based on Bee
Colony Algorithm.” In2015 Fifth International Conference on
hosts in order to save money and time on the Advances in Computing and Communications (ICACC), pp. 89-93.
network. IEEE, 2015.
[10] Magalhães, Deborah, Rodrigo N. Calheiros, RajkumarBuyya, and
3. In order to decrease the memory access time sent Danielo G. Gomes. "Workload modeling for resource usage analysis
and simulation in cloud computing." Computers & Electrical
back from the source host; the main problem with Engineering 47 (2015): 69-81.
the post copy method is that repetitive page fault [11] Cho, Keng-Mao, Pang-Wei Tsai, Chun-Wei Tsai, and Chu-Sing
detection needs to be handled. As a result, the post Yang. "A hybrid meta-heuristic algorithm for VM scheduling with
copy migration process will become more reliable. load balancing in cloud computing." Neural Computing and
Applications 26, no. 6 (2015): 1297-1309.
[12] Chen, Kun-Ting, Chien Chen, and Po-Hsiang Wang, “Network aware
V. CONCLUSION load-balancing via parallel VM migration for data centers,” In 2014
23rd International Conference on Computer Communication and
Realizing VM-based resource reservation requires VM Networks (ICCCN), pp. 1-8. IEEE, 2014.
[13] Komarasamy, , and VijayalakshmiMuthuswamy, “A Novel Approach
migration, and making effective reservations requires an for Dynamic Load Balancing with Effective Bin Packing and VM
awareness of its overhead. The virtualization environment, Reconfiguration in Cloud,” Indian Journal of Science and Technology
virtual machine migration, and ideal resource management 9, no. 11, 2016.
were just a few of the cloud computing elements that were [14] Sarker, TusherKumer, and Maolin Tang, “Performancedriven live
migration of multiple virtual machines in datacenters,” In Granular
highlighted. This study simulates the migratory process Computing (GrC), 2013 IEEE International Conference on, pp. 253-
based on a thorough investigation, yielding a number of 258. IEEE, 2013.
intriguing merits. Due to technical difficulties in the [15] Hu, Jinhua, JianhuaGu, Guofei Sun, and Tianhai Zhao. "A scheduling
migration process, it is still necessary to develop an ideal strategy on load balancing of virtual machine resources in cloud
computing environment." In Parallel Architectures, Algorithms and
migration model for virtual machines. From the survey, it is Programming (PAAP), 2010 Third International Symposium on, pp.
analyzed that sequential migration decreases the overhead 89-96. IEEE, 2010.
for the applications operating on the transferred VMs, while [16] M. Duggan, J. Duggan, E. Howley, and E. Barrett, “An autonomous
parallel migration increases the performance of migrating network aware vm migration strategy in cloud data centres,” in
Proceedingof the International Conference on Cloud and Autonomic
many VMs. Different migration options are available to Computing(ICCAC-2016), Sept 2016, pp. 24–32.
maximize the performance, which can suffer from VM [17] N. K. Sharma and G. R. M. Reddy, “A novel approach for
migration performance deterioration for a considerable multidimensional variable sized virtual machine allocation and
amount of time. In order to generalize the findings and migration at cloud data center,” in 2017 9th International Conference
on Communicatio Systems and Networks (COMSNETS), Jan 2017,
analyze it across many hardware platforms and pp. 383–384.
virtualization technologies; the

455

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.
 2022 5th International Conference on Contemporary Computing and Informatics (IC3I)

[18] M. Nelson, B. Lim, and G. Hutchines 2005. Fast transparent

Conference on Intelligent Engineering and Management (ICIEM),
migration for virtual machines. Proceedings of the USENIX Annual
London, UK, 2020, pp. 483-486, doi:
Technical Conference (USENIX’05), 391–394.
10.1109/ICIEM48762.2020.9160337.
[19] C. Christopher, F. Keir, H. Steven, H. Jacob Gorm, J. Eric, L. Christian,
[35] D. Singh, H. S. Garjan, S. V. Singh and G. Bhardhwaj, "A Novel
P. Ian, and W. Andrew 2005. Live migration of virtual machines. 2nd
Optimization Technique for Integrated Supply Chain Network in
conference on Symposium on Networked Systems Design &
Industries - A Technical Perspective," 2021 2nd International
Implementation- Volume 2: USENIX Association.
Conference on Intelligent Engineering and Management (ICIEM),
[20] H. Jin, L. Deng, S. Wu, X. H. Shi, and X. D. Pan 2009. Live Virtual
London, United Kingdom, 2021, pp. 550-554, doi:
Machine Migration with Adaptive Memory Compression. IEEE
10.1109/ICIEM51511.2021.9445392.
International Conference on Cluster Computing, 1-10.
[36] Gupta, A., Gupta, M., Chaturvedi, P. (2020). Investing Data with
[21] P. Svard, J. Tordsson, B. Hudzia, and E. Elmroth 2011. High
Machine Learning Using Python. In: Kapur, P.K., Singh, O., Khatri,
performance live migration through dynamic page transfer reordering
S.K., Verma, A.K. (eds) Strategic System Assurance and Business
and compression. 3rd IEEE International Conference on Cloud
Analytics. Asset Analytics. Springer, Singapore.
Computing Technology and Science, CloudCom, pp 542-548
https://doi.org/10.1007/978-981-15-3647-2_1
[22] B.Hu, Z. Lei, Y. Lei, D. Xu and J. Li 2011. A TimeSeries Based
[37] R. Gupta, G. Bhardwaj and G. Singh, "Employee Perception and
Precopy Approach for Live Migration of Virtual Machine. IEEE 17th
Behavioral Intention to Adopt BYOD in the Organizations," 2019
International Conference on Parallel and Distributed Systems ,pp 947-
International Conference on Automation, Computational and
952.
Technology Management (ICACTM), London, UK, 2019, pp. 73-78,
[23] K. Z. Ibrahim, S. Hofmeyr, C. Iancu, and E. Roman 2011. Optimized
doi: 10.1109/ICACTM.2019.8776815.
precopy live migration for memory intensive applications.
[38] Awasthi, S., Travieso-González, C. M., Sanyal, G., & Singh, D. K.
International Conference for High Performance Computing,
(Eds.). (2021). Artificial intelligence for a sustainable industry 4.0.
Networking, Storage and Analysis (SC),pp 1-11.
Springer International Publishing.
[24] R. H. Michael, D. Umesh, and G. Kartik 2009. Post-copy live
migration of virtual machines. ACM Special Interest Group on
Operating Systems, pp14-26.
[25] L. Weining and F. Tao 2009. Live migration of virtual machine based
on recovering system and CPU scheduling. 6th IEEE joint
International Information Technology and Artificial Intelligence
Conference,Piscataway, NJ, USA, 303- 208
[26] L. Haikun, J. Hai, L. Xiaofei, H. Liting, and Y. Chen, , 2009. Live
migration of virtual machine based on full system trace and replay.
18th ACM International Symposium on High performance
distributed computing, pp 101-110.
[27] Soramichi Akiyama, Takahiro Hirofuchi, Ryousei Takano and
Shinichi Honiden, “MiyakoDori: A Memory Reusing Mechanism for
Dynamic VM Consolidation”, 2012 IEEE Fifth International
Conference on Cloud Computing
[28] Arun Babu Nagarajan, Frank Mueller, Christian Engelmann, Stephen
L.Scott 2007. Proactive fault tolerance for HPC with Xen
Virtualization. Proceedings of the 21st annual international
conference on supercomputing.
[29] Kunwar, V., Agarwal, N., Rana, A., Pandey, J.P. (2018). Load
Balancing in Cloud—A Systematic Review. In: Aggarwal, V.,
Bhatnagar, V., Mishra, D. (eds) Big Data Analytics. Advances in
Intelligent Systems and Computing, vol 654. Springer, Singapore.
https://doi.org/10.1007/978-981-10-6620-7_56
[30] M. S. Raghavendra, P. Chawla and A. Rana, "A Survey of
Optimization Algorithms for Fog Computing Service Placement,"
2020 8th International Conference on Reliability, Infocom
Technologies and Optimization (Trends and Future Directions)
(ICRITO), Noida, India, 2020, pp. 259-262, doi:
10.1109/ICRITO48877.2020.9197885.
[31] Gupta, S., Rana, A., Kansal, V. (2020). Optimization in Wireless
Sensor Network Using Soft Computing. In: Raju, K., Govardhan, A.,
Rani, B., Sridevi, R., Murty, M. (eds) Proceedings of the Third
International Conference on Computational Intelligence and
Informatics . Advances in Intelligent Systems and Computing, vol
1090. Springer, Singapore. https://doi.org/10.1007/978-981-15-1480-
7_74
[32] G. Singh, P. Chaturvedi, A. Shrivastava and S. Vikram Singh, "Breast
Cancer Screening Using Machine Learning Models," 2022 3rd
International Conference on Intelligent Engineering and Management
(ICIEM), London, United Kingdom, 2022, pp. 961-967, doi:
10.1109/ICIEM54221.2022.9853047.
[33] G. Singh, G. Bhardwaj, S. V. Singh, P. Chaturvedi, V. Kumar and S.
Gupta, "Industry 4.0: The Industrial Revolution and Future Landscape
in Indian Market," 2021 International Conference on Technological
Advancements and Innovations (ICTAI), Tashkent, Uzbekistan, 2021,
pp. 500-505, doi: 10.1109/ICTAI53825.2021.9673154.
[34] V. Kumar, D. Singh, S. V. Singh and T. Anand, "The Role of
Converged Network in Disruptive Technology," 2020 International

456

Authorized licensed use limited to: SARDAR VALLABHBHAI NATIONAL INSTITUTE OF TECH. Downloaded on March 24,2023 at 10:34:05 UTC from IEEE Xplore.
Restrictions apply.