This document discusses load balancing in cloud computing. It begins by defining cloud computing and some of its key characteristics like broad network access, rapid elasticity, and pay-as-you-go pricing. It then discusses how load balancing can improve performance in distributed cloud environments by redistributing load, improving response times, and better utilizing resources. The document outlines different load balancing techniques like virtual machine migration and throttled load balancing using a load balancer, virtual machines, and a data center controller. It also proposes a trust and reliability based algorithm that prioritizes data centers for load balancing based on calculated trust values that consider factors like initialization time, machine performance, and fault rates.
11. Throttled load balancing
Data center
controller
Load balancer
Finds VM
Sends VM Id to DCC
Assign request to VM
Index table of VM &
state of VM
DCC notifies LB
12. Trust and Reliability based Algorithm
•Better distribution of load and improved QOS
•Trust value that is calculated by the trust model on basis of different
parameters are used to prioritize the datacenters and scheduling algorithms
to make use of these trust values ie. To schedule the request based on trust
values.
•Based on trust value load balancing is done for better selection of node for
better balancing of load.
•In cloud Iaas, we use load balancing algo to distribute the load, better
resource utilization and better QOS to the user.
•Based on trust values of the domain resources are assigned to the user
13. Trust based Algorithm
•Trust model are used to calculate the trust values for datacenters based on
the parameters which are as follows:
1. Initialization time : time taken to allocate resources requested and
deploy them
2. Machine Instruction Per Second(MIPS) : Number of instructions
completed per second
3. Fault rate: Number of faults in a period of time
•Trust values are calculated for each datacenters
•Updated after a fixed period of time
14. Whole cycle –
1. Initialization- trust value of data center and client is initialized
•data center is assigned the trust value on the basis of parameters and
updated periodically
•If the new data center is introduced, it is assigned with default trust value
•All the parameters are initialized
•Fault rate is initialized with default values for all datacenters
15. 2. Load Balancing
•Done using prioritized list of data centers and client trust
•Listing of trusted and untrusted datacenters is done
•Trusted node list consists of nodes having trust value greater than
threshold value in decreasing order ie. The first node has the highest trust
value
•Untrusted node list consists of nodes having trust value lesser than
threshold value in decreasing order
16. Working
•1. untrusted user
•When node gets overloaded, VM with lowest trust value is migrated
•Check selected VM is of public user or private user
•If public, node to which VM is to be migrated is selected from untrusted
list
•2. trusted user
17. 3. Trust Evolution
•Updating datacentre trust value periodically after a fixed interval of time
•Monitor behaviour of the datacentre on basis of fault rate and initialization
time
•Parameter value changes when load on data center increases/decreases
•Let datacenters Di and attributes Ai1(initialization time), Ai2(MIPS),
Ai3(Fault rate) and their corresponding trust value Ti
•Value to be updated-