Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Copy of Innovation Ecosystem
Transcript of Copy of Innovation Ecosystem
- Literature Review & Related Works
- Modeling and Simulation
- Future Work
- Acknowledgment The Simulation Model of
One : One (R1+R2) P(0,0)= F1 P(1,0) + F2 P(0,1)
(F1+R2) P(1,0) = R1 P(0,0) +F1 P(1,1)
(F2+R1) P(0,1) = R2 P(0,0) +F1 P(1,1)
(F2+F1) P(1,1) = R2 P(1,0) +R1 P(0,1) The Analytical Model of One : One The availability of the private service subsystem The analytical model was verified using the simulator results.
Two Both models results shown to be similar
6 factors were considered,
MTTR and MTTF for each resource. Evaluation Numerical Results The system was decomposed into two subsystems:
the private service: (resource + network (LAN))
the public service: (resource + network (Internet))
A (system) = A(private) + (U(private) * A(public) | network(public is Up)) (1) The Analytical Model of One : One Numerical Results 1:1 One : One Service Protection Model in a Hybrid Cloud Computing Architecture Background Research Question Thanks for Listening
firstname.lastname@example.org • This thesis aims to answer the following research question:
What is the impact significance of different factors on the service availability of the protection models in a hybrid cloud computing architecture.? Dr. Ayman Fayoumi Outlines What is Cloud Computing? Thesis Objectives The main objective of this thesis is to evaluate the availability of the service for the protection models in a hybrid cloud computing architecture. Aprivate = MTTFprivate/ MTTFprivate+MTTRprivate The availability of the public service subsystem: This work was published in the proceeding of
IEEE ISTT 2012. The Simulation Model of One : N Calculating the availability for the public resource The Analytical Model of One : N Evaluation of the availability of this model were done for ten times each with different number of competitors.
Matlab was used to obtain the availability of the analytical model.
The Statistics of the simulation model were printed by the simulator. Evaluation Numerical Results The Analytical Model of One : N 1:N One : N Service Protection Model in a Hybrid Cloud Computing Architecture The Simulation Model of M : N Calculating the availability for the public resource : The Analytical Model of M : N Evaluation of the availability of this model were done for six times with an assumption of 10 competitors competing on 4 public cloud resources each with :
different values for the MTTF and MTTR of the private resource
different values for the MTTF and MTTR of the network resource
Increasing no of public resources
Increasing no of competitors
Matlab was used to obtain the availability of the analytical model. The Statistics of the simulation model were printed by the simulator. Evaluation A (system) = A (private) + ( U (private) | ( network (public) is Up AND
Num (busy_public )< M) )
The availability of the private service subsystem:
A (private) = MTTF(private) / MTTF(private) + MTTR(private)
The availability of the public service subsystem:
Calculating the availability of the network
A (network) = MTTF(network) / MTTF(network)+ MTTR(network) The Analytical Model of M : N Numerical Results M:N P(busy) = P(Free) = 1 - P(Busy) The author makes derivations and calculate the ratio of the change for each factor (differential impact) to represent the change in the availability per one unit of the tested factor.
The numerical results of this derivation show that increasing the MTTF of the private resource has the higher impact on the availability then the impact of the other factors can be classified as follows from the highest impact to the lowest:
increasing MTTF of the private resource
decreasing MTTR of the private resource
increasing MTTF of the network resource
decreasing MTTR of the network resource
increasing number of resources
decreasing number of competitors Literature Review & Related Works Factors IMPACT SHARED SCALED Conclusion Future Work An analysis and evaluation of the service performance in cloud computing is needed. This will help to complete the image of the performability in cloud computing.
The research could be extended to investigate and study the service reliability of cloud computing.
More factors may affect the service availability in cloud computing (software or database failures).
The first model can be repeated considering wait requests while the protecting resource is busy. (timeout failure)
The second model can be repeated considering interruption, and the protection goes to serve another request before ending the current request.
The third model can be repeated considering that the protection resources in a “down” state. Publications 1. Wazzan, M. and Fayoumi, A., (2012) "A Survey of Researches on the Challenge Issues in Cloud Computing," INFORMATION: An International Interdisciplinary Journal, vol 15, 2012 –ISI Journal
(an invitaion )
2. Wazzan, M. and Fayoumi, A., (2012) "Service Availability Evaluation for a Protection Model in Hybrid Cloud Computing Architecture, " in the proceeding of 1st IEEE International Symposium on Telecommunication Technologies (ISTT2012), 26-28 November
3. Wazzan, M. and Fayoumi, A. Service Availability Evaluation for 1:N Protection Model in Hybrid Cloud Computing Architecture. (to be sent to Publication)
4. Wazzan, M. and Fayoumi, A. Service Availability Evaluation for M:N Protection Model in Hybrid Cloud Computing Architecture. (sent to "Elsevier performance evaluation" an ISI international journal ) This thesis presents three studies to evaluate the availability of the service in protection models in a hybrid cloud computing architecture.
It uses two referenced techniques
The Markov model
The Discrete event simulation
It examines the impact of the following factors:
MTTF , MTTR, No of Competitors, No of protection resources.
The results shows the impact of each factors on the availability.
A classification depending on the impact of each factor on the system availability was given. Acknowledgment Dr. Ayman Fayoumi
To KACST for financially supporting this research A (system) = A(private) + ( U(private) * P(public (Free) |
(network (public) is Up AND Public Resource is availabe))
The availability of the private service subsystem :
A(private) = MTTF(private) / MTTF(private) + MTTR(private)
The availability of the public service subsystem :
Calculating the availability of the network
A(network) = MTTF(network) / MTTF(network)+ MTTR(network) M : N Service Protection Model in a Hybrid Cloud Computing Architecture Conclusion, Future Work & Publications Which Factor has the Highest Impact on the Service Availability? The Analytical Model of One : One Discrete event simulation is a mathematical model employing to estimate the system behavior using a computer program. Availability On The Performability
of Cloud Computing Modeling and Simulation Availability means whether the service performs
its design function or not at a given instant of time. Availability is one of the measures of the
performability of cloud computing. Protection techniques can be used to preserve the cloud services. Service protection techniques can be one-to-one
(1:1) protection , one-to-N (1:N)
protection, or M-to-N (M:N) protection Markov model is a mathematical technique significantly used in calculating availability
of various systems. It generates a set of equations that need a
solution to complete the analysis. Discrete Event Simulation Building the Simulator An algorithm was developed for the simulator. A simulator was built step by step It explores the system behavior by using a random-numbers generator. It tracks the states of the system according to the occurrence of the events in a discrete instance of time during the simulation. A linked list and pointers was used to schedule events. Each event scheduled according to its TTBP. Then it removed from the list after processing. The clock of the simulator is advanced to the process time of the processed event The cumulative statistics of the used counters are updated and Steady state was checked. This work is ready to sent to publication This work was sent to "Elsevier Performance evaluation" ISI International Journal