J. Salehi, J. Kurose and D. Towsley. "Further Results in Affinity-Based Scheduling of Parallel Networking". TRUMCS -1995-046, U. Massachusetts, May 1995. (available via ftp from gaia.cs.umass.edu in pub/Sale95:Further.ps.Z) Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Networking Support For High-Performance Servers - Nahum (1997) (Correct)
....take advantage of the machines full capabilities. One way to improve performance in the network protocol subsystem to exploit the availability of multiple processors in the host. The use of parallelism in network protocol processing has recently become an active area of research in both academia [5, 12, 20, 21, 22, 35, 46, 47, 48, 59, 62, 63, 64, 69, 70, 71, 72, 74, 75, 76, 77, 87, 88, 89, 95, 100, 106, 107, 108, 109, 111, 112, 116, 117, 124, 125] and industry [18, 37, 42, 45, 49, 68, 90, 94, 110, 120] Many approaches to parallelism in network protocols have been proposed. We provide a brief taxonomy of parallelism in protocols here; more detailed surveys can be found in [12, 48] In general, we attempt to classify approaches by the unit ....
Salehi, J. D., Kurose, J. F., and Towsley, D. Further results in affinity-based scheduling of parallel networking. Technical Report UM-CS-1995-046, Department of Computer Science, University of Massachusetts, Amherst, MA, May 1995.
The Performance Impact of Scheduling for Cache Affinity.. - Salehi, Kurose, Towsley (1995) (2 citations) Self-citation (Salehi Kurose Towsley) (Correct)
....predominate is that typically in real environments most packets are small (e.g. 5, 10] 6 . We do point out how Figures 10 and 11 (below) can be interpreted to evaluate the impact of data touching operations on affinity based scheduling under Locking and IPS, respectively. Further, in [18] we explicitly incorporate into our results the overhead of copying packet data. 5.1 Affinity scheduling under Locking Figures 6 and 7 explore the performance of affinitybased scheduling under Locking, plotting mean packet delay as a function of packet arrival rate. To isolate the marginal ....
....be managed MRU except under high arrival rate, when Wired Streams scheduling performs better. Under IPS, independent stacks should be wired to processors except under low arrival rate, when MRU processor scheduling performs better. We are currently pursuing several extensions to this work [18], including i) evaluating affinity based scheduling of send side UDP IP FDDI processing; ii) examining the performance of affinity based scheduling as a function of stream burstiness and source locality, as captured by the Packet Train model of [9] iii) exploring under IPS the impact of varying ....
J. Salehi, J. Kurose and D. Towsley. "Further Results in Affinity-Based Scheduling of Parallel Networking". TRUMCS -1995-046, U. Massachusetts, May 1995. (available via ftp from gaia.cs.umass.edu in pub/Sale95:Further.ps.Z)
The Effectiveness of Affinity-Based Scheduling in.. - Salehi, Kurose, Towsley (1996) (14 citations) Self-citation (Salehi Kurose Towsley) (Correct)
....stream specific data structures written during a send operation, as a first step toward developing send side PLP and CLP implementations. In contrast to the receive side path, send side UDP IP FDDI processing references only read only protocol specific and stream specific data structures (see [29] for discussion) Thus, there is no need to explicitly enable parallelism through software locks (PLP) or through data structure replication (CLP) and the unparallelized implementation immediately supports concurrency. Thus, there is no PLP CLP distinction with regard to send side UDP ....
....code affinity scheduling. The impacts of stream burstiness and source locality on the performance of affinity based scheduling of receive side CLP, as well of send side processing, are similar. The qualitative behaviors and relationships of the various scheduling policies are maintained. See [29] for the complete set of graphs. Overall, apart from the queueing effect seen under stream affinity scheduling, it is somewhat surprising that stream burstiness has such a small impact on the qualitative performance of the affinity based scheduling policies. After all, it is widely known that the ....
James Salehi, James Kurose, and Don Towsley. Further results in affinity-based scheduling of parallel networking. Technical Report UM-CS-1995-046, University of Massachusetts, Amherst, May 1995.
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