1. © Performance Analysis And Comparison Of Interrupt-Handling Schemes In Gigabit Networks Salah, K; El-Badawi, K; Haidari, F ELSEVIER SCIENCE BV, COMPUTER COMMUNICATIONS; pp: 3425-3441; Vol: 30 King Fahd University of Petroleum & Minerals http://www.kfupm.edu.sa Summary Interrupt processing can be a major bottleneck in the end-to-end performance of Gigabit networks. The performance of Gigabit network end hosts or servers can be severely degraded due to interrupt overhead caused by heavy incoming traffic. In particular, excessive latency and significant degradation in system throughput can be encountered. Also, user applications may livelock as the CPU power gets mostly consumed by interrupt handling and protocol processing. A number of interrupt- handling schemes has been proposed and employed to mitigate the interrupt overhead and improve OS performance. Among the most popular interrupt-handling schemes are normal interruption, polling, interrupt coalescing, and disabling and enabling of interrupts. In previous work, we presented a preliminary analytical study and models of normal interruption and interrupt coalescing. In this article, we extend our analysis and modeling to include polling and the scheme of interrupt disabling and enabling. For polling, we study both pure (or FreeBSD-style) polling and Linux NAPI polling. The performances for all these schemes are compared using both mathematical analysis and discrete-event simulation. The performance is studied in terms of three key performance indicators: throughput, system latency, and the residual CPU bandwidth available for user applications. As opposed to our previous work, we consider not only Poisson traffic, but also bursty traffic with empirical packet size distribution. Our analysis and simulation work gives insight into predicting the system performance and behavior when employing a certain interrupt-handling scheme. It is concluded that no single interrupt-handling scheme outperforms all other schemes Copyright: King Fahd University of Petroleum & Minerals; http://www.kfupm.edu.sa

2. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. © under all traffic conditions. Based on obtained results, we propose and discuss a novel hybrid scheme of interrupt disabling-enabling and pure polling in order to attain peak performance under low and heavy traffic loads. (c) 2007 Elsevier B.V. All rights reserved. References: ASHFORD COMPUTER CON *3COM CORP, GIG SERV NETW INT CA *ALT WEBSYSTEMS IN, JUMB FRAM ALOUF S, 2001, P IEEE INFOCOM 2001, V2, P1045 ARON M, 2000, ACM T COMPUT SYST, V18, P197 BHOEDJANG RAF, 1998, IEEE COMPUT, V31, P53 BIANCO A, 2005, P 3 INT WORKSH QOS M, P353 BROWN A, 1997, P 1997 ACM SIGMETRIC, P214 BRUN O, 2000, J APPL PROBAB, V37, P1092 BRUSTOLONI J, 1996, P 2 USENIX S OP SYST, P277 CHASE JS, 2001, IEEE COMMUN MAG, V39, P68 CLAFFY KC, 1998, P INET 1998 GEN SWIT CLARK D, 2002, IEEE COMMUNICATION M, V40, P94 CROVELLA ME, 1997, P WINT SIM C, P1005 DERI L, 2004, P 4 INT SYST ADM NET DITTA Z, 1997, P IEEE INFOCOM 1997, P179 DOVROLIS C, 2001, ACM OPERATING SYSTEM, V35, P50 DRUSCHEL P, 1996, COMMUN ACM, V39, P41 DRUSCHEL P, 1996, P 2 S OP SYST DES IM, P261 DUNIGAN T, ORNL OPTERON EVALUAT DUNKELS A, DESIGN IMPLEMENTATIO FENG W, 2000, P SC2000 DALL TEX US FOONG A, 2003, IEEE S PERF SYST SOF, P70 GALLATIN A, 1999, TRAPEZE IP TCP IP NE INDIRESAN A, 1998, RECEIVE LIVELOCK ELI KARAM MJ, 2002, COMPUT NETW, V40, P711 KENG H, 1996, P USENIX 1996 ANN TE KIM I, 2001, P 5 INT C HIGH PERF KOCHETKOV K, INTEL PRO 1000 DESKT LAW A, 1991, SIMULATION MODELING LELAND WE, 1994, IEEE ACM T NETWORK, V2, P1 MALIK S, 2004, P PERF MOD EV HET NE MAQUELIN O, 1996, P 23 ANN INT S COMP, P178 MCKUSICK M, 1996, DESIGN IMPLEMENTATIO MCVOY L, 1996, P USENIX 1996 ANN TE, P279 MOGUL JC, 1997, ACM T COMPUT SYST, V15, P217 MORRIS R, 2000, ACM T COMPUT SYST, V8, P263 Copyright: King Fahd University of Petroleum & Minerals; http://www.kfupm.edu.sa

3. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. © PARK K, 2002, SELF SIMILAR NETWORK PAXSON V, 1995, IEEE ACM T NETWORK, V3, P226 PRASAD R, 2004, P PASS ACT MEAS PAM RAMAKRISHNAN KK, 1993, IEEE J SEL AREA COMM, V11, P203 RUBINI A, 2001, LINUX DEVICE DRIVERS SALAH K, 2003, 28 IEEE LOC COMP NET, P498 SALAH K, 2003, IEEE GLOBECOM 2003 S, P3953 SALAH K, 2005, INT J COMMUN SYST, V18, P501, DOI 10.1002/dac.727 SALAH K, 2007, INT J COMPUTER SYSTE, V22 SALAH K, 2007, INT J ELECT COMM AEU, V61, P215 SALAH K, 2007, P IEEE 2U INT C ADV, P1000 SALIM JH, 2001, P 5 ANN LIN SHOWC C, P165 SHIVAN P, 2001, P SC2001 DENV COL US SINHRA A, 2004, P USENIX TECHN C BOS, P213 TAQQU MS, 1997, ACM SIGCOMM COMPUTER, V24, P5 TRAW C, 1993, IEEE NETWORK, V7, P44 TRAW CBS, 1993, IEEE J SEL AREA COMM, V11, P240 WANG CL, 2003, ACM T MODELING COMPU, V13, P62 WHITE KP, 1997, SIMULATION, V69, P323 WILLINGER W, 1995, P ACM SIGCOMM 95, P100 ZEC M, 2002, P 10 SOFTCOM For pre-prints please write to: salah@kfupm.edu.sa; elbadawi@kfup-m.edu.sa; fahd@kfupm.edu.sa Copyright: King Fahd University of Petroleum & Minerals; http://www.kfupm.edu.sa