Final

LOAD BALANCING IN CLOUD
COMPUTING
Guide: Team Members
P.Kanagaraju K.B.Manojkumar
SS.Gauthaamman
G.Deepak
OBJECTIVE
 As request demand expands, contemporary
servers can be appended to the resource pool,
where the load balancer will straightway start
sending traffic to the new server.
INTRODUCTION
 Load balancing techniques
NAÏVE SHARED BASED APPROACH OF
LOAD BALANCING
•It is a combination of Throttled and

Current Execution algorithms that
equally spread.
• Proposed algorithm overcomes the
drawbacks of throttled and equally
spread current execution algorithm.
LOAD BALANCING
Equally spread current
Throttled Algorithm
Execution Algorithm
LOAD BALANCING
A COMPREHENSIVE STUDY REPORT ON LOAD
BALANCING TECHNIQUES IN CLOUD
COMPUTING
 Load is a measure of the amount of work that a
computation system performs which can be
classified as CPU load, network load, memory
capacity and storage capacity.
 To Achieve high user satisfaction, resource
utilization by ensuring efficient and fair
allocation of computing resources.
AIMS AND CHALLENGES
 Increases
 Availability of services
 User satisfaction
 Resource utilization
 Improve performance
 Decrease
 Execution and waiting time
 Overhead Associated
 Throughput
 Scalability
 Fault tolerance
 Point of failure
ALGORITHMS
 Round Robin Algorithm
 Work load distributions are equal but the job
processing time for different processes are not same.
So at any point of time some nodes may be heavily
loaded and others remain idle.
 Weighted Round Robin
 Does not consider the advanced load balancing
requirements such as processing times for each
individual requests .
 Dynamic Round Robin
 This algorithm mainly works for reducing the power
consumption of physical machine.
 Randomized
 The algorithm is very simple but job may require
long waiting time before being serviced
ALGORITHMS
 Equally Spread Current Execution (ESCE) Algorithm:
 It is not fault tolerant and has the problem of single point of
failure.
 Throttled
 The current load on the VM is not considered which can in turn
increase the response time of a task.
 Modified Throttled
 It gives better response time.But in index table the state of some
VM may change during the allocation of next request due to de-
allocation of some tasks. So it is not always beneficial to start
searching from the next to already assigned VM.
 Central Load Balancer
 The algorithm efficiently balances in heterogeneous environment
but it suffers from bottleneck as all the requests will come to
central load balancer. Its based on priority of VMs which is
calculated in a static way and is not updated during job
allocation.
ALGORITHMS
 Active Monitoring Load Balancing (AMLB)
Algorithm
 During allocation of VM only importance is given on
the current load of VM, its processing power is not
taken into consideration. So the waiting time of some
jobs may increase violating the QoS requirement.
 VM-Assign Load Balancing Algorithm
 This algorithm will not use the VM if it is already
allocated in the last round
 Weighted Active Monitoring Load Balancing
Algorithm
 Removes shortcomings of Active Monitoring Load
Balancing Algorithm by not only considering the load
but also the processing power of available VMs.
ALGORITHM
 Biased radndom sampling

 Min-Min scheduling algorithm
 Load balance Improved Min-Min scheduling

algorithm
 User-Priority Awared Load Balance Improved
Min-Min Scheduling Algorithm (PA-LBIMM)
 Opportunistic Load Balancing (OLB)
 Honey bee raging algorithm
 Active Clustering
LOAD BALANCING
 Load balancing has two meanings:
 first, it puts a large number of concurrent accesses
or data traffic to multiple nodes respectively to
reduce the time users waiting for response.
 second, it put the calculation from a single heavy
load to the multiple nodes to improve the resource
utilization of each node.
 Classified as
 Static algorithm
 Dynamic algorithm
TRADITIONAL COMPUTING VS CLOUD
COMPUTING
 Traditional
 It mainly schedules process or task so the
granularity and the transferred data is small.
 Cloud computing
 The scheduling target is VM resources so the
granularity is large and the transferred data is large
as well.
 The time of scheduling algorithm can almost be
neglected.
COMPARISON
LOAD BALANCING
 Classsified
 Sender Initiated
 Receiver Initiated
 Symmetric
 Static
 RRA
 Min-Min Algorithm
 Min-Max Algorithm
 Dynamic
STRUCTURE OF LOAD BALANCING

Final

Uploaded by

Copyright:

Available Formats

Final

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Final

Uploaded by

Copyright:

Available Formats

LOAD BALANCING IN CLOUD

•It is a combination of Throttled and

 Biased radndom sampling

 Load balance Improved Min-Min scheduling

 Honey bee raging algorithm

You might also like