V. Vuppala, L. M. Ni; "Design of A Scalable IP Router"; Proc. IEEE Hot Interconnects 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....interfaces connect router nodes to a high speed interconnect, whereas the external interfaces connect Panama to the rest of the network. DMA to optimize throughput by avoiding multiple trips between the I O bus and memory. This techniques is also used in Panama. The Scalable IP Router Project [16] proposes to use a cluster architecture to speed up complicated IP route computation such as OSPF, rather than IP packet forwarding or processing. In Panama, PC clustering is viewed as a means to scale the performance of both packet forwarding and computation paths, and the challenge is to develop ....

V. Vuppala, L. M. Ni; "Design of A Scalable IP Router"; Proc. IEEE Hot Interconnects 1997.

....new deadlocks, andf urther global reasoning will not be necessary. 1 Introduction Wormh8k routing (WHR) h, recently beenbrough t into new applications,such ashk h speed LANs and SANs (i.e. Autonet [1] Myrinet [2] and Servernet [3] andh as been proposed as th e internal fabric of switch es [4,5]. Th e latter is interesting forhk h ly scalable switch es. One could consider using a k ary n cube [6] as th internal topologywhz building such a switch ,butth result will h ve an unfortunate property. Unlike a simple crossbar router ,th WHR switch may block,thU is,a packet may h ve to wait for ....

Vibhavasu Vuppala and Lionel M. Ni. Design of A Scalable IP Router. In Hot In tercon cts V, 1997.

....improve the performance of the most expensive part of forwarding, the longest prefix match into the forwarding table. Our solution falls into the category of switching techniques but can be used in conjunction with System Area Networks (SAN) 16] 6] to provide a scalable high throughput router [33]. 2.1 Switching Techniques IP Switching 1 [20] 21] was one of the first switching techniques. It dynamically chooses between IP routing and ATM switching, depending on the characteristics of the network traffic. It identifies a flow in IP traffic and assigns it a virtual circuit in the ATM ....

....Path VNP For virtual ports N 1 P 2 and N 1 P 3 , node N 2 is the next hop node and the network segment between N 1 and N 2 is the next hop segment. As mentioned earlier, the next hop segment can be a LAN, WAN or SAN segment. A SAN segment can be used to develop a highly scalable router as in [33] or ATM fabric can be used to have a campus or ISP wide distributed router . Certain network interfaces like LAN interfaces connect to a network but others like point to point or loop back interfaces connect to a node. So the destination of a VNP can be a network or a node. If the destination is ....

V. Vuppala and L. M. Ni, "Design of a Scalable IP Router", Hot Interconnects V, Aug 1997.

....overview of the limitations of PC hardware and possible suggestions for performance optimizations. 24] proposes the use of peer to peer DMA to optimize throughput by avoiding multiple trips between the I O bus and memory. This techniques is also used in Xrouter. The Scalable IP Router Project [28] proposes to use a cluster architecture to speed up complicated IP route computation such as in OSPF, rather than IP packet forwarding or processing. In Xrouter, PC clustering is viewed as a means to scale the performance of both packet forwarding and computation paths, and the challenge is to ....

V. Vuppala, L. M. Ni; "Design of A Scalable IP Router"; Proc. IEEE Hot Interconnects 1997.

....in IP switching [3] that sets up an ATM connection for long term IP connections. The work was similar to Suez in that the link layer transmission are based on fixed sized data units, which greatly simplifies real time packet scheduling. The scalable IP router project in Michigan State University [2] is based on the same hardware software platform as Suez. However, that project s focus is on parallelizing route computation to address route instability problems rather than fast routing table lookup or QoS provisions. Similar to IP switching, Suez employs off the shelf high performance ....

....each cache block contains at most 8 HAC entries. Finally, assume one quarter of the L1 cache, i.e. 128 out of 512 cache sets, is reserved for HAC. 3 0 0 0 0 bits [11, 5] bits [31, 12] OutPort Tag 32 bytes 384 sets 512 sets 32 bits L1 Cache Translation Address Cache Lookup bits [4, 2] Address Address Physical ignored bits [1, 0] Software Compare Destination IP Address (DA) Virtual HAC Figure 1: The data flow of HAC lookup. Here we assume that L1 cache is direct mapped with 512 32 byte cache sets, among which 128 cache sets are reserved for HAC. Each HAC entry is 4 ....

Vuppala, V. and Ni, L., "Design of a Scalable IP Router," Hot Interconnect, Stanford, CA., August 1997.

....overview of the limitations of PC hardware and possible suggestions for performance optimizations. 24] proposes the use of peer to peer DMA to optimize throughput by avoiding multiple trips from the I O bus to memory and back. This techniques is also used in Suez. The Scalable IP Router Project [28] proposes to use a cluster architecture to speed up complicated IP route computation such as in OSPF, rather than IP packet forwarding or processing. In Suez, PC clustering is viewed as a means to scale the performance of both packet forwarding and computation paths, and the challenge is to ....

V. Vuppala, L. M. Ni; "Design of A Scalable IP Router"; Proc. IEEE Hot Interconnects 1997.

No context found.

Vibhavasu Vuppala and Lionel M. Ni, "Design of A Scalable IP Router", Proceedings of the 1997 IEEE Hot Interconnect, Aug. 1997.

No context found.

Vibhavasu Vuppala and Lionel M. Ni, "Design of A Scalable IP Router", Proceedings of the 1997 IEEE Hot Interconnect, Aug. 1997.

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