A queuing model for the distributed queue dual bus protocol is presented and analyzed. A queue bu... more A queuing model for the distributed queue dual bus protocol is presented and analyzed. A queue buffer of length one and a single priority level are assumed. The state of a node is defined by the number of requests in the distributed queue before and after generation of a packet. The node average waiting time and average throughput are analyzed
ATM is the target switching and multiplexing technology for implementation of B-ISDN. However, ex... more ATM is the target switching and multiplexing technology for implementation of B-ISDN. However, existing ATM software systems are deficient in many respects, such as QoS-based routing. Noting that the route of a virtual circuit remains fixed throughout a session, it is clear that route selection will have long-term effects on network congestion, especially for long sessions. Hence, it is important
n this chapter, we propose a new approach for content delivery services by meshing together the b... more n this chapter, we propose a new approach for content delivery services by meshing together the best of grid computing and peer-to-peer (P2P) computing. The goal is to design a secure, reliable, and scalable system for efficient and fast delivery of content. The system consists of a combination of nondedicated servers and peers to provide the proposed service. We describe the challenges of designing such a system and discuss possible solutions and trade-offs. We detail the necessary interlacing of grid and P2P feature to achieve the goal. We present a prototype that is built based on the proposed approach.
A queueing model for the DQDB protocol with a single priority level is presented and analyzed. Th... more A queueing model for the DQDB protocol with a single priority level is presented and analyzed. The state of a node is defined by the number of requests in the distributed queue prior to, and post generation of a segment. ft is shown that the number of states that a node can be in, is finite and position dependent. Moreover, it is shown that the request rate and the network traffic that a node perceives at any particdm moment are highly dependent on its state. An iterathe technique is presented to calculate the request rate and network trafiic for all the nodes in a DQDB network by carefully modePing the relationship between adjacent nodes. The node average waiting time and average throughput are analyzed as fuuctions of the node position and state in the network. Our results illustrate the effect of the various parameters on the waiting time and throughput. Comparisons with simulation and existing analytical results are presented.
... those VMs. The Job VM Host-ing Environment is the equivalent of the User Hosting Environment ... more ... those VMs. The Job VM Host-ing Environment is the equivalent of the User Hosting Environment in GT3, with the modification that it runs in a Linux VM different from the Manager VM Hosting Environment. The overall sequence ...
Enterprises are under great pressure to be nimble and adjust quickly to ever shifting business pr... more Enterprises are under great pressure to be nimble and adjust quickly to ever shifting business priorities and constantly evolving technologies. This is usually seen as a driver to adopt Cloud technologies, however, the existing large investments in on premise IT infrastructure and data residency constraints combined with decreasing IT budgets precludes a fast and immediate migration to public or private clouds, hence the realization that Hybrid IT environment are here to stay for a while, making the efficient management of hybrid IT environments extremely critical. The pressure of decreasing IT budgets also precludes the high cost of investment in traditional tools and systems to manage those environments. This has resulted in the rise of an on-demand software-as-a-service (SaaS) solutions to manage these hybrid clouds. However, these SaaS solutions are typically delivered as island solutions and lack some of the characteristics that SaaS solutions in this space should exhibit in order to provide an efficient Enterprise IT management eco-system. In this paper, we first present the set of properties that SaaS solutions must exhibit in order to become part of the eco-system and describe the architecture of a platform where those SaaS solutions can operate and integrate with each other. We present the challenges of building such a platform, and how our architecture addresses these challenges. We showcase how we are leveraging this platform to transform a large service management provider's IT services to fit this new paradigm.
A queuing model for the distributed queue dual bus protocol is presented and analyzed. A queue bu... more A queuing model for the distributed queue dual bus protocol is presented and analyzed. A queue buffer of length one and a single priority level are assumed. The state of a node is defined by the number of requests in the distributed queue before and after generation of a packet. The node average waiting time and average throughput are analyzed
ATM is the target switching and multiplexing technology for implementation of B-ISDN. However, ex... more ATM is the target switching and multiplexing technology for implementation of B-ISDN. However, existing ATM software systems are deficient in many respects, such as QoS-based routing. Noting that the route of a virtual circuit remains fixed throughout a session, it is clear that route selection will have long-term effects on network congestion, especially for long sessions. Hence, it is important
n this chapter, we propose a new approach for content delivery services by meshing together the b... more n this chapter, we propose a new approach for content delivery services by meshing together the best of grid computing and peer-to-peer (P2P) computing. The goal is to design a secure, reliable, and scalable system for efficient and fast delivery of content. The system consists of a combination of nondedicated servers and peers to provide the proposed service. We describe the challenges of designing such a system and discuss possible solutions and trade-offs. We detail the necessary interlacing of grid and P2P feature to achieve the goal. We present a prototype that is built based on the proposed approach.
A queueing model for the DQDB protocol with a single priority level is presented and analyzed. Th... more A queueing model for the DQDB protocol with a single priority level is presented and analyzed. The state of a node is defined by the number of requests in the distributed queue prior to, and post generation of a segment. ft is shown that the number of states that a node can be in, is finite and position dependent. Moreover, it is shown that the request rate and the network traffic that a node perceives at any particdm moment are highly dependent on its state. An iterathe technique is presented to calculate the request rate and network trafiic for all the nodes in a DQDB network by carefully modePing the relationship between adjacent nodes. The node average waiting time and average throughput are analyzed as fuuctions of the node position and state in the network. Our results illustrate the effect of the various parameters on the waiting time and throughput. Comparisons with simulation and existing analytical results are presented.
... those VMs. The Job VM Host-ing Environment is the equivalent of the User Hosting Environment ... more ... those VMs. The Job VM Host-ing Environment is the equivalent of the User Hosting Environment in GT3, with the modification that it runs in a Linux VM different from the Manager VM Hosting Environment. The overall sequence ...
Enterprises are under great pressure to be nimble and adjust quickly to ever shifting business pr... more Enterprises are under great pressure to be nimble and adjust quickly to ever shifting business priorities and constantly evolving technologies. This is usually seen as a driver to adopt Cloud technologies, however, the existing large investments in on premise IT infrastructure and data residency constraints combined with decreasing IT budgets precludes a fast and immediate migration to public or private clouds, hence the realization that Hybrid IT environment are here to stay for a while, making the efficient management of hybrid IT environments extremely critical. The pressure of decreasing IT budgets also precludes the high cost of investment in traditional tools and systems to manage those environments. This has resulted in the rise of an on-demand software-as-a-service (SaaS) solutions to manage these hybrid clouds. However, these SaaS solutions are typically delivered as island solutions and lack some of the characteristics that SaaS solutions in this space should exhibit in order to provide an efficient Enterprise IT management eco-system. In this paper, we first present the set of properties that SaaS solutions must exhibit in order to become part of the eco-system and describe the architecture of a platform where those SaaS solutions can operate and integrate with each other. We present the challenges of building such a platform, and how our architecture addresses these challenges. We showcase how we are leveraging this platform to transform a large service management provider's IT services to fit this new paradigm.
Uploads
Papers by Dikran Meliksetian