| A. Neogi, A. Raniwala, T.Chiueh. \Phoenix: a low-power fault-tolerant real-time network-attached storage device." In Proceedings of ACM Multimedia 1999. |
....information and meta data are replicated to improve fault tolerance. Petal s high scalability and fault tolerance come from this totally distributed storage architecture. However, none of the above mentioned systems tried to solve the quality of service problem. To address the QoS issues, Phoenix [36, 42] is a fault tolerant network attached storage device with quality of service guarantees. It behaves like one single storage node in a shared clustered storage system. Kernel is made QoS conscious and can accept real time streams. The admission control and scheduling algorithms are based on ....
Neogi, A., Raniwala, A., and cker Chiueh, T. Phoenix: a low-power fault-tolerant real-time network-attached storage device. In ACM Multimedia (1999), pp. 447-456.
....clients. It is not a complete QoS solution bacouse it only takes into account bandwidth resource which does not necessiraly reflect the real system resource consumption. Several research efforts focused on QoS guarantee for specific type of resource, such as CPU[25, 6, 16, 29, 8, 19, 17] disk[20, 9, 28], and on shared network link[21, 24, 14] Almost all of these systems can be abstracted into the same implementation frmework as described in this thesis. In particular, one can show the fluid fair queuing model[27] provides the theoretical bases for all these resource schedulers. However, ....
{phoenix} A. Neogi, A. Raniwala, T.Chiueh. "Phoenix: a low-power fault-tolerant real-time network-attached storage device." In Proceedings of ACM Multimedia 1999, Orlando, FL, October 1999.
....those from Alteon [2] Cisco [10] Foundry [15] at most support only content aware and load balancing request dispatching to a Web server cluster, but not Web server QoS. Several research e orts focused on QoS guarantee for speci c type of resource, such as CPU [25, 6, 16, 29, 8, 19, 17] disk [20, 9, 28], and on shared network link [21, 24, 14] Almost all of these systems can be abstracted into the same implementation framework as described in this paper. In particular, one can show the uid fair queuing model [27] provides the theoretical basis for all these resource schedulers. However, ....
A. Neogi, A. Raniwala, T.Chiueh. \Phoenix: a low-power fault-tolerant real-time network-attached storage device." In Proceedings of ACM Multimedia 1999.
....Alteon [2] Cisco [10] Foundry [15] at most support only content aware and loadbalancing request dispatching to a Web server cluster, but not Web server QoS. Several research efforts focused on QoS guarantee for specific type of resource, such as CPU [25] 6] 16] 29] 8] 19] 17] disk [20], 9] 28] and on shared network link [21] 24] 14] Almost all of these systems can be abstracted into the same implementation framework as described in this paper. In particular, one can show the fluid fair queuing model [27] provides the theoretical basis for all these resource schedulers. ....
A. Neogi, A. Raniwala, T.Chiueh. "Phoenix: a low-power fault-tolerant real-time network-attached storage device." In Proceedings of ACM Multimedia 1999.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC