{"title":"建模和模拟操作系统抖动的影响","authors":"Elder Vicente, Rivalino Matias","doi":"10.5555/2675983.2676250","DOIUrl":null,"url":null,"abstract":"The phenomenon of operating system (OS) Jitter has been investigated and considered a critical factor in high-performance computing. In this paper we model and simulate the effects of different sources of OS Jitter in the Linux operating system. We adopt the design of experiments approach to conduct experiments statistically planned. Our simulation models corroborate the results obtained experimentally. We conclude that the OS Jitter has a higher impact when the number of the computational phases is high, for any number of computing nodes from 1 to 500. We also observed that in Linux, the highest OS Jitter impacts are caused by managing the shared processor cache and network interrupts, where the second shows the highest sensitivity with respect to the cluster size.","PeriodicalId":223717,"journal":{"name":"2013 Winter Simulations Conference (WSC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2013-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":"{\"title\":\"Modeling and simulating the effects of OS Jitter\",\"authors\":\"Elder Vicente, Rivalino Matias\",\"doi\":\"10.5555/2675983.2676250\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The phenomenon of operating system (OS) Jitter has been investigated and considered a critical factor in high-performance computing. In this paper we model and simulate the effects of different sources of OS Jitter in the Linux operating system. We adopt the design of experiments approach to conduct experiments statistically planned. Our simulation models corroborate the results obtained experimentally. We conclude that the OS Jitter has a higher impact when the number of the computational phases is high, for any number of computing nodes from 1 to 500. We also observed that in Linux, the highest OS Jitter impacts are caused by managing the shared processor cache and network interrupts, where the second shows the highest sensitivity with respect to the cluster size.\",\"PeriodicalId\":223717,\"journal\":{\"name\":\"2013 Winter Simulations Conference (WSC)\",\"volume\":\"27 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2013-12-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"6\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2013 Winter Simulations Conference (WSC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.5555/2675983.2676250\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 Winter Simulations Conference (WSC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.5555/2675983.2676250","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
The phenomenon of operating system (OS) Jitter has been investigated and considered a critical factor in high-performance computing. In this paper we model and simulate the effects of different sources of OS Jitter in the Linux operating system. We adopt the design of experiments approach to conduct experiments statistically planned. Our simulation models corroborate the results obtained experimentally. We conclude that the OS Jitter has a higher impact when the number of the computational phases is high, for any number of computing nodes from 1 to 500. We also observed that in Linux, the highest OS Jitter impacts are caused by managing the shared processor cache and network interrupts, where the second shows the highest sensitivity with respect to the cluster size.