SlideShare a Scribd company logo
1 of 24
PROXIMAL VIRTUAL
NETWORK
EMBEDDING
BASED ON MULTI-
DIMENSIONAL LOAD
BALANCING
CSE Department, NITK
Surathkal
TEAM
• Himanshu Pandey
(232CS011)
• Md Talib (232CS017)
• Piyus Prabhanjans
(232CS024)
Idea Behind The Paper:
TO ADDRESS THE CHALLENGES OF
EFFICIENT VIRTUAL NETWORK
EMBEDDING (VNE) IN DATA
CENTERS.
NEW VNE ALGORITHMS TO BE
DESIGNED FOR FULL
UTILIZATION OF REGULAR AND
SYMMETRIC NATURE OF DATA
CENTER NETWORKS.
TO PREVENT RESOURCE WASTE
AND IMPROVE OVERALL RESOURCE
UTILIZATION.
Objectives:
Efficient Virtual Network Embedding (VNE): for efficient allocation
of physical resources in data centers for flexible resource
allocation
Resource Utilization Improvement: the goal of the algorithm to
increase resource utilization in data centers by optimizing resource
allocation and reducing transmission hops.
Load Balancing Enhancement: the algorithm aims to achieve load
balancing not only for single-dimensional resources but also for
multi-dimensional resources to prevent resource waste and improve
overall performance.
Motivation behind this Paper:
TO AVOID LOAD IMBALANCE, AS
IT IS PROMINENT IN DATA
CENTERS DUE TO POOLING OF
DATA CENTER RESOURCES,
MULTITENANCY AND
HETEROGENEITY
CHARACTERISTICS.
MULTIPLE PREVIOUS STUDIES
ONLY DID LOAD BALANCING IN
SINGLE DIMENSION WHICH MAY
LEAD TO UNDERUTILIZATION OR
OVERUTILIZATION OF SPECIFIC
RESOURCES, LIMITING OVERALL
EFFICIENCY.
TRADITIONAL VNE ALGORITHMS
ARE TAILORED FOR GENERIC
INTERNET TOPOLOGIES AND MAY
NOT FULLY EXPLOIT THE
REGULAR AND SYMMETRIC NATURE
OF DATA CENTER NETWORKS.
What is VNE?
• Virtual Network Embedding (VNE) refers to the
process of mapping virtual networks onto
physical infrastructure in a virtualized
environment, such as data centers or cloud
computing platforms. In VNE, the goal is to
efficiently allocate computing, storage, and
network resources to virtual networks while
ensuring optimal performance, resource
utilization, and cost-effectiveness. This
involves mapping the logical topology of virtual
networks, including virtual nodes and links,
onto the physical topology of the underlying
infrastructure, which may consist of servers,
switches, and links. VNE algorithms aim to
address challenges such as resource allocation,
topology mapping, load balancing, and
scalability to optimize the mapping of virtual
networks onto physical infrastructure and ensure
What is Load Balancing?
• Load balancing refers to the distribution of workloads
across multiple resources or servers to ensure optimal
utilization, performance, and reliability. In computing,
load balancing involves evenly distributing incoming
network traffic, requests, or computational tasks among
available servers, nodes, or resources. The primary
objectives of load balancing are to prevent overloading
of individual resources, avoid bottlenecks, maximize
resource utilization, and maintain system stability and
responsiveness. Load balancing mechanisms dynamically
allocate incoming requests or tasks to different
resources based on factors such as current resource
utilization, network conditions, and predefined policies
or algorithms.
• This ensures that no single resource is overwhelmed,
thereby enhancing system scalability, fault tolerance,
and overall performance. Load balancing is essential in
various computing environments, including data centers,
cloud computing platforms, and distributed systems, to
efficiently manage and optimize resource usage and
ensure seamless operation of applications and services.
Why Load Balancing is
required in VNE?
● Resource Utilization: Load balancing ensures that the computational and
network resources in the data center are utilized optimally. By evenly
distributing virtual network requests across physical resources, load
balancing prevents certain nodes or links from being underutilized while
others are overloaded.
● Performance Optimization: Efficient load balancing improves the overall
performance of the network by minimizing congestion and reducing response
times. It ensures that no single server or link becomes a bottleneck,
which can degrade the performance of virtual networks and applications.
● Fault Tolerance: Load balancing enhances the fault tolerance of the system
by redistributing traffic away from failed or degraded resources. In case
of hardware failures or network congestion, load balancing algorithms
redirect traffic to healthy resources, ensuring uninterrupted service
availability.
● Scalability: As the number of virtual network requests increases, load
balancing helps scale the system by dynamically allocating resources to
accommodate the growing demand. It allows the network to handle
fluctuations in traffic load efficiently without compromising performance
or stability.
● Cost Efficiency: Effective load balancing can lead to cost savings by
optimizing resource usage and reducing the need for additional hardware or
infrastructure. By maximizing the utilization of existing resources,
organizations can minimize operational expenses associated with data
● Load Imbalance Problem of VNE in Data Centers
Illustrative Example:
● Figure 1 is presented to illustrate examples of VNE
scenarios.
● Fig. 1(a) depicts two virtual network requests.
● Fig. 1(b) and Fig. 1(c) demonstrate the mapping
results of two different VNE schemes within a pod in
a data center based on fat-tree topology.
● The gray area in each server represents the current
computing resource occupancy status, and the
numbers beside each access link represent the
current bandwidth occupancy status.
Observati
ons about
Paper:
• The mapping result of the green VN
request using embedding without
Multi-Dimensional Load Balancing
(MLB) leads to:
• Imbalance of computing loads
within the pod.
• Computing resource exhaustion
of one server, making it
impossible to map all the
virtual nodes of the orange VN
request within the same pod.
• Cross-pod node mapping
resulting in long-path link
mapping.
• There's an increase in the number
of hops due to the imbalance.
• We can observe that, the mapping
result of the green VN request
using embedding without MLB leads
to the imbalance of computing loads
within the pod and the computing
resource exhaustion of one server,
which makes it impossible to map
all the virtual nodes of the orange
VN request within the same pod.
About
Implementations:
Solution based on Proximal VNE
based on multi-dimensional load
balancing
Since this is a combinatorial optimization problem, which is
NP-hard. Therefore, we design a heuristic algorithm to get a
suboptimal solution.
Solution is based upon:
a) Node Mapping:
Node mapping in virtual network embedding (VNE) refers to the process
of assigning virtual nodes (representing virtual machines or other
network components) to physical nodes (servers or switches) within a
data center network. This mapping is essential for translating the logical
structure of virtual networks onto the physical infrastructure.
During node mapping, various factors are considered to ensure efficient
utilization of resources and optimal performance:
1. Resource Availability
2. Topology Constraints
3. Load Balancing
4. Fault Tolerance:
For Node Mapping:
Function in code: node_map(substrate, virtual, req_no)
Purpose: Node mapping determines which substrate nodes
will host the virtual nodes of a given virtual network
request (VNR). The goal is to find suitable substrate
nodes that can accommodate the resource requirements of
the virtual nodes.
Approach: The function iterates over the virtual nodes and
attempts to find a suitable substrate node for each
virtual node based on available resources. It employs a
worst-fit strategy to select the substrate node with the
maximum available capacity that meets the requirements of
the virtual node.
Algorithm
for Node
Mapping:
• Algorithm:
• Sort the substrate nodes in descending order of
available capacity (e.g., computational
resources, bandwidth).
• For each virtual node:
• Find the substrate node with the
maximum available capacity that satisfies
the resource requirements of the virtual
node.
• Assign the virtual node to the selected
substrate node.
• Update the available capacity of the
selected substrate node.
b) Link Mapping:
Link mapping in virtual network embedding (VNE)
involves determining the paths or routes that virtual
links will take within the physical network infrastructure.
Once virtual nodes are mapped to physical nodes, link
mapping establishes the connections between these
nodes by selecting the appropriate physical links.
Factors are considered to ensure efficient utilization of
resources and optimal performance during link
mapping :
1. Bandwidth Availability
2. Latency
3. Load Balancing
4. Resilience and Fault Tolerance
5. Topology Constraints
6. Quality of Service (QoS) Requirements
7. Resource Optimization Objectives
For Link Mapping:
Function in code: edge_map(substrate, virtual, req_no,
req_map, vne_list)
Purpose: Edge mapping determines how virtual edges will
be mapped onto substrate paths while ensuring that
bandwidth constraints are met. It aims to establish
substrate paths between the mapped virtual nodes.
Approach: The function iterates over the virtual edges
and attempts to find suitable substrate paths that can
accommodate the bandwidth requirements of the virtual
edges. It utilizes a modified breadth-first search (BFS)
algorithm to find substrate paths between mapped virtual
nodes.
Algorithm
for Link
Mapping:
• Use a modified BFS algorithm
to find a path between the
mapped substrate nodes of the
virtual edge.
• Ensure that the bandwidth
constraints of the virtual
edge are satisfied along the
substrate path.
• Update the bandwidth
availability of the substrate
edges along the selected
path.
• The worst-fit concept
is applied when selecting
substrate paths with
sufficient bandwidth to
accommodate the virtual
edges.
• The algorithm prioritizes
selecting substrate paths
with the maximum available
bandwidth that meets the
OUTPUT : Visualization and Comparisons
• Case-1
• Case-2
• Case-3
• All Cases Comparision
Average Bandwidth for Three Cases
• Average Node for Three Cases
• Average Path for Three Cases
THANK YOU

More Related Content

Similar to Ccpro_Presentationfddgdddcfbbjjdssvhbvxf11.pptx

Paper sharing_resource optimization scheduling and allocation for hierarchica...
Paper sharing_resource optimization scheduling and allocation for hierarchica...Paper sharing_resource optimization scheduling and allocation for hierarchica...
Paper sharing_resource optimization scheduling and allocation for hierarchica...YOU SHENG CHEN
 
Iisrt komathi krishna (networks)
Iisrt komathi krishna (networks)Iisrt komathi krishna (networks)
Iisrt komathi krishna (networks)IISRT
 
On the Optimal Allocation of VirtualResources in Cloud Compu.docx
On the Optimal Allocation of VirtualResources in Cloud Compu.docxOn the Optimal Allocation of VirtualResources in Cloud Compu.docx
On the Optimal Allocation of VirtualResources in Cloud Compu.docxhopeaustin33688
 
Resource optimization-based network selection model for heterogeneous wireles...
Resource optimization-based network selection model for heterogeneous wireles...Resource optimization-based network selection model for heterogeneous wireles...
Resource optimization-based network selection model for heterogeneous wireles...IAESIJAI
 
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...An Efficient Distributed Control Law for Load Balancing in Content Delivery N...
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...IJMER
 
A novel cost-based replica server placement for optimal service quality in ...
  A novel cost-based replica server placement for optimal service quality in ...  A novel cost-based replica server placement for optimal service quality in ...
A novel cost-based replica server placement for optimal service quality in ...IJECEIAES
 
Cost aware cooperative resource provisioning
Cost aware cooperative resource provisioningCost aware cooperative resource provisioning
Cost aware cooperative resource provisioningIMPULSE_TECHNOLOGY
 
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...iosrjce
 
dynamic resource allocation using virtual machines for cloud computing enviro...
dynamic resource allocation using virtual machines for cloud computing enviro...dynamic resource allocation using virtual machines for cloud computing enviro...
dynamic resource allocation using virtual machines for cloud computing enviro...Kumar Goud
 
Combining efficiency, fidelity, and flexibility in resource information services
Combining efficiency, fidelity, and flexibility in resource information servicesCombining efficiency, fidelity, and flexibility in resource information services
Combining efficiency, fidelity, and flexibility in resource information servicesieeepondy
 
A Distributed Control Law for Load Balancing in Content Delivery Networks
A Distributed Control Law for Load Balancing in Content Delivery NetworksA Distributed Control Law for Load Balancing in Content Delivery Networks
A Distributed Control Law for Load Balancing in Content Delivery NetworksSruthi Kamal
 
IEEE Networking 2016 Title and Abstract
IEEE Networking 2016 Title and AbstractIEEE Networking 2016 Title and Abstract
IEEE Networking 2016 Title and Abstracttsysglobalsolutions
 
Service oriented cloud architecture for improved performance of smart grid ap...
Service oriented cloud architecture for improved performance of smart grid ap...Service oriented cloud architecture for improved performance of smart grid ap...
Service oriented cloud architecture for improved performance of smart grid ap...eSAT Journals
 
Service oriented cloud architecture for improved
Service oriented cloud architecture for improvedService oriented cloud architecture for improved
Service oriented cloud architecture for improvedeSAT Publishing House
 
Survey on Dynamic Resource Allocation Strategy in Cloud Computing Environment
Survey on Dynamic Resource Allocation Strategy in Cloud Computing EnvironmentSurvey on Dynamic Resource Allocation Strategy in Cloud Computing Environment
Survey on Dynamic Resource Allocation Strategy in Cloud Computing EnvironmentEditor IJCATR
 

Similar to Ccpro_Presentationfddgdddcfbbjjdssvhbvxf11.pptx (20)

Paper sharing_resource optimization scheduling and allocation for hierarchica...
Paper sharing_resource optimization scheduling and allocation for hierarchica...Paper sharing_resource optimization scheduling and allocation for hierarchica...
Paper sharing_resource optimization scheduling and allocation for hierarchica...
 
Iisrt komathi krishna (networks)
Iisrt komathi krishna (networks)Iisrt komathi krishna (networks)
Iisrt komathi krishna (networks)
 
On the Optimal Allocation of VirtualResources in Cloud Compu.docx
On the Optimal Allocation of VirtualResources in Cloud Compu.docxOn the Optimal Allocation of VirtualResources in Cloud Compu.docx
On the Optimal Allocation of VirtualResources in Cloud Compu.docx
 
Resource optimization-based network selection model for heterogeneous wireles...
Resource optimization-based network selection model for heterogeneous wireles...Resource optimization-based network selection model for heterogeneous wireles...
Resource optimization-based network selection model for heterogeneous wireles...
 
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...An Efficient Distributed Control Law for Load Balancing in Content Delivery N...
An Efficient Distributed Control Law for Load Balancing in Content Delivery N...
 
A novel cost-based replica server placement for optimal service quality in ...
  A novel cost-based replica server placement for optimal service quality in ...  A novel cost-based replica server placement for optimal service quality in ...
A novel cost-based replica server placement for optimal service quality in ...
 
Cost aware cooperative resource provisioning
Cost aware cooperative resource provisioningCost aware cooperative resource provisioning
Cost aware cooperative resource provisioning
 
17 51-1-pb
17 51-1-pb17 51-1-pb
17 51-1-pb
 
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...
The Grouping of Files in Allocation of Job Using Server Scheduling In Load Ba...
 
J017367075
J017367075J017367075
J017367075
 
dynamic resource allocation using virtual machines for cloud computing enviro...
dynamic resource allocation using virtual machines for cloud computing enviro...dynamic resource allocation using virtual machines for cloud computing enviro...
dynamic resource allocation using virtual machines for cloud computing enviro...
 
unit3 part1.pptx
unit3 part1.pptxunit3 part1.pptx
unit3 part1.pptx
 
Resource management
Resource managementResource management
Resource management
 
Combining efficiency, fidelity, and flexibility in resource information services
Combining efficiency, fidelity, and flexibility in resource information servicesCombining efficiency, fidelity, and flexibility in resource information services
Combining efficiency, fidelity, and flexibility in resource information services
 
A Distributed Control Law for Load Balancing in Content Delivery Networks
A Distributed Control Law for Load Balancing in Content Delivery NetworksA Distributed Control Law for Load Balancing in Content Delivery Networks
A Distributed Control Law for Load Balancing in Content Delivery Networks
 
2012an20
2012an202012an20
2012an20
 
IEEE Networking 2016 Title and Abstract
IEEE Networking 2016 Title and AbstractIEEE Networking 2016 Title and Abstract
IEEE Networking 2016 Title and Abstract
 
Service oriented cloud architecture for improved performance of smart grid ap...
Service oriented cloud architecture for improved performance of smart grid ap...Service oriented cloud architecture for improved performance of smart grid ap...
Service oriented cloud architecture for improved performance of smart grid ap...
 
Service oriented cloud architecture for improved
Service oriented cloud architecture for improvedService oriented cloud architecture for improved
Service oriented cloud architecture for improved
 
Survey on Dynamic Resource Allocation Strategy in Cloud Computing Environment
Survey on Dynamic Resource Allocation Strategy in Cloud Computing EnvironmentSurvey on Dynamic Resource Allocation Strategy in Cloud Computing Environment
Survey on Dynamic Resource Allocation Strategy in Cloud Computing Environment
 

Recently uploaded

Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2T.D. Shashikala
 
The Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxThe Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxMANASINANDKISHORDEOR
 
Diploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfDiploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfJNTUA
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsMathias Magdowski
 
Augmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxAugmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxMustafa Ahmed
 
"United Nations Park" Site Visit Report.
"United Nations Park" Site  Visit Report."United Nations Park" Site  Visit Report.
"United Nations Park" Site Visit Report.MdManikurRahman
 
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisSeismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisDr.Costas Sachpazis
 
Introduction to Artificial Intelligence and History of AI
Introduction to Artificial Intelligence and History of AIIntroduction to Artificial Intelligence and History of AI
Introduction to Artificial Intelligence and History of AISheetal Jain
 
Autodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxAutodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxMustafa Ahmed
 
Low Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookLow Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookPeterJack13
 
Dynamo Scripts for Task IDs and Space Naming.pptx
Dynamo Scripts for Task IDs and Space Naming.pptxDynamo Scripts for Task IDs and Space Naming.pptx
Dynamo Scripts for Task IDs and Space Naming.pptxMustafa Ahmed
 
CLOUD COMPUTING SERVICES - Cloud Reference Modal
CLOUD COMPUTING SERVICES - Cloud Reference ModalCLOUD COMPUTING SERVICES - Cloud Reference Modal
CLOUD COMPUTING SERVICES - Cloud Reference ModalSwarnaSLcse
 
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfMadan Karki
 
Worksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxWorksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxMustafa Ahmed
 
handbook on reinforce concrete and detailing
handbook on reinforce concrete and detailinghandbook on reinforce concrete and detailing
handbook on reinforce concrete and detailingAshishSingh1301
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdfAlexander Litvinenko
 
Performance enhancement of machine learning algorithm for breast cancer diagn...
Performance enhancement of machine learning algorithm for breast cancer diagn...Performance enhancement of machine learning algorithm for breast cancer diagn...
Performance enhancement of machine learning algorithm for breast cancer diagn...IJECEIAES
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashidFaiyazSheikh
 
Piping and instrumentation diagram p.pdf
Piping and instrumentation diagram p.pdfPiping and instrumentation diagram p.pdf
Piping and instrumentation diagram p.pdfAshrafRagab14
 
Artificial Intelligence in due diligence
Artificial Intelligence in due diligenceArtificial Intelligence in due diligence
Artificial Intelligence in due diligencemahaffeycheryld
 

Recently uploaded (20)

Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2
 
The Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptxThe Entity-Relationship Model(ER Diagram).pptx
The Entity-Relationship Model(ER Diagram).pptx
 
Diploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdfDiploma Engineering Drawing Qp-2024 Ece .pdf
Diploma Engineering Drawing Qp-2024 Ece .pdf
 
Filters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility ApplicationsFilters for Electromagnetic Compatibility Applications
Filters for Electromagnetic Compatibility Applications
 
Augmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptxAugmented Reality (AR) with Augin Software.pptx
Augmented Reality (AR) with Augin Software.pptx
 
"United Nations Park" Site Visit Report.
"United Nations Park" Site  Visit Report."United Nations Park" Site  Visit Report.
"United Nations Park" Site Visit Report.
 
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas SachpazisSeismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
Seismic Hazard Assessment Software in Python by Prof. Dr. Costas Sachpazis
 
Introduction to Artificial Intelligence and History of AI
Introduction to Artificial Intelligence and History of AIIntroduction to Artificial Intelligence and History of AI
Introduction to Artificial Intelligence and History of AI
 
Autodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptxAutodesk Construction Cloud (Autodesk Build).pptx
Autodesk Construction Cloud (Autodesk Build).pptx
 
Low Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s HandbookLow Altitude Air Defense (LAAD) Gunner’s Handbook
Low Altitude Air Defense (LAAD) Gunner’s Handbook
 
Dynamo Scripts for Task IDs and Space Naming.pptx
Dynamo Scripts for Task IDs and Space Naming.pptxDynamo Scripts for Task IDs and Space Naming.pptx
Dynamo Scripts for Task IDs and Space Naming.pptx
 
CLOUD COMPUTING SERVICES - Cloud Reference Modal
CLOUD COMPUTING SERVICES - Cloud Reference ModalCLOUD COMPUTING SERVICES - Cloud Reference Modal
CLOUD COMPUTING SERVICES - Cloud Reference Modal
 
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdfALCOHOL PRODUCTION- Beer Brewing Process.pdf
ALCOHOL PRODUCTION- Beer Brewing Process.pdf
 
Worksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxWorksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptx
 
handbook on reinforce concrete and detailing
handbook on reinforce concrete and detailinghandbook on reinforce concrete and detailing
handbook on reinforce concrete and detailing
 
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdflitvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
litvinenko_Henry_Intrusion_Hong-Kong_2024.pdf
 
Performance enhancement of machine learning algorithm for breast cancer diagn...
Performance enhancement of machine learning algorithm for breast cancer diagn...Performance enhancement of machine learning algorithm for breast cancer diagn...
Performance enhancement of machine learning algorithm for breast cancer diagn...
 
Raashid final report on Embedded Systems
Raashid final report on Embedded SystemsRaashid final report on Embedded Systems
Raashid final report on Embedded Systems
 
Piping and instrumentation diagram p.pdf
Piping and instrumentation diagram p.pdfPiping and instrumentation diagram p.pdf
Piping and instrumentation diagram p.pdf
 
Artificial Intelligence in due diligence
Artificial Intelligence in due diligenceArtificial Intelligence in due diligence
Artificial Intelligence in due diligence
 

Ccpro_Presentationfddgdddcfbbjjdssvhbvxf11.pptx

  • 1. PROXIMAL VIRTUAL NETWORK EMBEDDING BASED ON MULTI- DIMENSIONAL LOAD BALANCING CSE Department, NITK Surathkal
  • 2. TEAM • Himanshu Pandey (232CS011) • Md Talib (232CS017) • Piyus Prabhanjans (232CS024)
  • 3. Idea Behind The Paper: TO ADDRESS THE CHALLENGES OF EFFICIENT VIRTUAL NETWORK EMBEDDING (VNE) IN DATA CENTERS. NEW VNE ALGORITHMS TO BE DESIGNED FOR FULL UTILIZATION OF REGULAR AND SYMMETRIC NATURE OF DATA CENTER NETWORKS. TO PREVENT RESOURCE WASTE AND IMPROVE OVERALL RESOURCE UTILIZATION.
  • 4. Objectives: Efficient Virtual Network Embedding (VNE): for efficient allocation of physical resources in data centers for flexible resource allocation Resource Utilization Improvement: the goal of the algorithm to increase resource utilization in data centers by optimizing resource allocation and reducing transmission hops. Load Balancing Enhancement: the algorithm aims to achieve load balancing not only for single-dimensional resources but also for multi-dimensional resources to prevent resource waste and improve overall performance.
  • 5. Motivation behind this Paper: TO AVOID LOAD IMBALANCE, AS IT IS PROMINENT IN DATA CENTERS DUE TO POOLING OF DATA CENTER RESOURCES, MULTITENANCY AND HETEROGENEITY CHARACTERISTICS. MULTIPLE PREVIOUS STUDIES ONLY DID LOAD BALANCING IN SINGLE DIMENSION WHICH MAY LEAD TO UNDERUTILIZATION OR OVERUTILIZATION OF SPECIFIC RESOURCES, LIMITING OVERALL EFFICIENCY. TRADITIONAL VNE ALGORITHMS ARE TAILORED FOR GENERIC INTERNET TOPOLOGIES AND MAY NOT FULLY EXPLOIT THE REGULAR AND SYMMETRIC NATURE OF DATA CENTER NETWORKS.
  • 6. What is VNE? • Virtual Network Embedding (VNE) refers to the process of mapping virtual networks onto physical infrastructure in a virtualized environment, such as data centers or cloud computing platforms. In VNE, the goal is to efficiently allocate computing, storage, and network resources to virtual networks while ensuring optimal performance, resource utilization, and cost-effectiveness. This involves mapping the logical topology of virtual networks, including virtual nodes and links, onto the physical topology of the underlying infrastructure, which may consist of servers, switches, and links. VNE algorithms aim to address challenges such as resource allocation, topology mapping, load balancing, and scalability to optimize the mapping of virtual networks onto physical infrastructure and ensure
  • 7. What is Load Balancing? • Load balancing refers to the distribution of workloads across multiple resources or servers to ensure optimal utilization, performance, and reliability. In computing, load balancing involves evenly distributing incoming network traffic, requests, or computational tasks among available servers, nodes, or resources. The primary objectives of load balancing are to prevent overloading of individual resources, avoid bottlenecks, maximize resource utilization, and maintain system stability and responsiveness. Load balancing mechanisms dynamically allocate incoming requests or tasks to different resources based on factors such as current resource utilization, network conditions, and predefined policies or algorithms. • This ensures that no single resource is overwhelmed, thereby enhancing system scalability, fault tolerance, and overall performance. Load balancing is essential in various computing environments, including data centers, cloud computing platforms, and distributed systems, to efficiently manage and optimize resource usage and ensure seamless operation of applications and services.
  • 8. Why Load Balancing is required in VNE? ● Resource Utilization: Load balancing ensures that the computational and network resources in the data center are utilized optimally. By evenly distributing virtual network requests across physical resources, load balancing prevents certain nodes or links from being underutilized while others are overloaded. ● Performance Optimization: Efficient load balancing improves the overall performance of the network by minimizing congestion and reducing response times. It ensures that no single server or link becomes a bottleneck, which can degrade the performance of virtual networks and applications. ● Fault Tolerance: Load balancing enhances the fault tolerance of the system by redistributing traffic away from failed or degraded resources. In case of hardware failures or network congestion, load balancing algorithms redirect traffic to healthy resources, ensuring uninterrupted service availability. ● Scalability: As the number of virtual network requests increases, load balancing helps scale the system by dynamically allocating resources to accommodate the growing demand. It allows the network to handle fluctuations in traffic load efficiently without compromising performance or stability. ● Cost Efficiency: Effective load balancing can lead to cost savings by optimizing resource usage and reducing the need for additional hardware or infrastructure. By maximizing the utilization of existing resources, organizations can minimize operational expenses associated with data
  • 9. ● Load Imbalance Problem of VNE in Data Centers Illustrative Example: ● Figure 1 is presented to illustrate examples of VNE scenarios. ● Fig. 1(a) depicts two virtual network requests. ● Fig. 1(b) and Fig. 1(c) demonstrate the mapping results of two different VNE schemes within a pod in a data center based on fat-tree topology. ● The gray area in each server represents the current computing resource occupancy status, and the numbers beside each access link represent the current bandwidth occupancy status.
  • 10. Observati ons about Paper: • The mapping result of the green VN request using embedding without Multi-Dimensional Load Balancing (MLB) leads to: • Imbalance of computing loads within the pod. • Computing resource exhaustion of one server, making it impossible to map all the virtual nodes of the orange VN request within the same pod. • Cross-pod node mapping resulting in long-path link mapping. • There's an increase in the number of hops due to the imbalance. • We can observe that, the mapping result of the green VN request using embedding without MLB leads to the imbalance of computing loads within the pod and the computing resource exhaustion of one server, which makes it impossible to map all the virtual nodes of the orange VN request within the same pod.
  • 12. Solution based on Proximal VNE based on multi-dimensional load balancing Since this is a combinatorial optimization problem, which is NP-hard. Therefore, we design a heuristic algorithm to get a suboptimal solution. Solution is based upon: a) Node Mapping: Node mapping in virtual network embedding (VNE) refers to the process of assigning virtual nodes (representing virtual machines or other network components) to physical nodes (servers or switches) within a data center network. This mapping is essential for translating the logical structure of virtual networks onto the physical infrastructure. During node mapping, various factors are considered to ensure efficient utilization of resources and optimal performance: 1. Resource Availability 2. Topology Constraints 3. Load Balancing 4. Fault Tolerance:
  • 13. For Node Mapping: Function in code: node_map(substrate, virtual, req_no) Purpose: Node mapping determines which substrate nodes will host the virtual nodes of a given virtual network request (VNR). The goal is to find suitable substrate nodes that can accommodate the resource requirements of the virtual nodes. Approach: The function iterates over the virtual nodes and attempts to find a suitable substrate node for each virtual node based on available resources. It employs a worst-fit strategy to select the substrate node with the maximum available capacity that meets the requirements of the virtual node.
  • 14. Algorithm for Node Mapping: • Algorithm: • Sort the substrate nodes in descending order of available capacity (e.g., computational resources, bandwidth). • For each virtual node: • Find the substrate node with the maximum available capacity that satisfies the resource requirements of the virtual node. • Assign the virtual node to the selected substrate node. • Update the available capacity of the selected substrate node.
  • 15. b) Link Mapping: Link mapping in virtual network embedding (VNE) involves determining the paths or routes that virtual links will take within the physical network infrastructure. Once virtual nodes are mapped to physical nodes, link mapping establishes the connections between these nodes by selecting the appropriate physical links. Factors are considered to ensure efficient utilization of resources and optimal performance during link mapping : 1. Bandwidth Availability 2. Latency 3. Load Balancing 4. Resilience and Fault Tolerance 5. Topology Constraints 6. Quality of Service (QoS) Requirements 7. Resource Optimization Objectives
  • 16. For Link Mapping: Function in code: edge_map(substrate, virtual, req_no, req_map, vne_list) Purpose: Edge mapping determines how virtual edges will be mapped onto substrate paths while ensuring that bandwidth constraints are met. It aims to establish substrate paths between the mapped virtual nodes. Approach: The function iterates over the virtual edges and attempts to find suitable substrate paths that can accommodate the bandwidth requirements of the virtual edges. It utilizes a modified breadth-first search (BFS) algorithm to find substrate paths between mapped virtual nodes.
  • 17. Algorithm for Link Mapping: • Use a modified BFS algorithm to find a path between the mapped substrate nodes of the virtual edge. • Ensure that the bandwidth constraints of the virtual edge are satisfied along the substrate path. • Update the bandwidth availability of the substrate edges along the selected path. • The worst-fit concept is applied when selecting substrate paths with sufficient bandwidth to accommodate the virtual edges. • The algorithm prioritizes selecting substrate paths with the maximum available bandwidth that meets the
  • 18. OUTPUT : Visualization and Comparisons • Case-1
  • 21. • All Cases Comparision Average Bandwidth for Three Cases
  • 22. • Average Node for Three Cases
  • 23. • Average Path for Three Cases