{"title":"动态SDN控制平面有多灵活?","authors":"Mu He, A. Basta, Andreas Blenk, W. Kellerer","doi":"10.1109/INFCOMW.2017.8116460","DOIUrl":null,"url":null,"abstract":"In Software Defined Networking (SDN), for performance reasons the control plane consists of multiple SDN controllers that provide a logically centralized network control. To react to traffic changes, the control plane may adapt to improve the performance of the whole network. This adaptation includes the migration of controllers, i.e. the change of their placement, as well as the change of switch to controller assignments. We call such adaptive network control a dynamic SDN control plane. Whereas most state of the art focuses on an optimal placement of controllers considering static traffic demands, we draw the attention to the need for a dynamic control plane and the performance of the adaption process itself. The involved controller migration and switch re-assignments significantly affect the control plane performance. Moreover, in this work we evaluate the flexibility of a dynamic SDN control plane with respect to the number of controllers. In particular, we refer to a flexibility metric that takes the time into account that is needed for successful migrations to handle traffic changes. Our evaluations are based on detailed models of controller placement and migration time. The results confirm that the migration time is a critical parameter and reveal that in case only a short migration time is allowed, the number of controllers does not bring more flexibility. For relaxed migration times, a larger number of controllers leads to more flexibility as expected. The results also show that a dynamic adaptation to traffic changes, can provide an improvement of up to 60% over a non-dynamic SDN control plane.","PeriodicalId":306731,"journal":{"name":"2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)","volume":"305 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"16","resultStr":"{\"title\":\"How flexible is dynamic SDN control plane?\",\"authors\":\"Mu He, A. Basta, Andreas Blenk, W. Kellerer\",\"doi\":\"10.1109/INFCOMW.2017.8116460\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In Software Defined Networking (SDN), for performance reasons the control plane consists of multiple SDN controllers that provide a logically centralized network control. To react to traffic changes, the control plane may adapt to improve the performance of the whole network. This adaptation includes the migration of controllers, i.e. the change of their placement, as well as the change of switch to controller assignments. We call such adaptive network control a dynamic SDN control plane. Whereas most state of the art focuses on an optimal placement of controllers considering static traffic demands, we draw the attention to the need for a dynamic control plane and the performance of the adaption process itself. The involved controller migration and switch re-assignments significantly affect the control plane performance. Moreover, in this work we evaluate the flexibility of a dynamic SDN control plane with respect to the number of controllers. In particular, we refer to a flexibility metric that takes the time into account that is needed for successful migrations to handle traffic changes. Our evaluations are based on detailed models of controller placement and migration time. The results confirm that the migration time is a critical parameter and reveal that in case only a short migration time is allowed, the number of controllers does not bring more flexibility. For relaxed migration times, a larger number of controllers leads to more flexibility as expected. The results also show that a dynamic adaptation to traffic changes, can provide an improvement of up to 60% over a non-dynamic SDN control plane.\",\"PeriodicalId\":306731,\"journal\":{\"name\":\"2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)\",\"volume\":\"305 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2017-05-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"16\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/INFCOMW.2017.8116460\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/INFCOMW.2017.8116460","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 16
摘要
在SDN (Software Defined Networking)中,出于性能考虑,控制平面由多个SDN控制器组成,这些控制器提供逻辑上集中的网络控制。为了应对流量的变化,控制平面可以做出相应的调整,从而提高整个网络的性能。这种自适应包括控制器的迁移,即它们位置的变化,以及切换到控制器分配的变化。我们称这种自适应网络控制为动态SDN控制平面。考虑到静态交通需求,大多数技术都集中在控制器的最佳位置上,我们提请注意对动态控制平面的需求和自适应过程本身的性能。所涉及的控制器迁移和交换机重分配对控制平面性能影响较大。此外,在这项工作中,我们根据控制器的数量评估了动态SDN控制平面的灵活性。特别地,我们引用了一个灵活性度量,它考虑了成功迁移处理流量更改所需的时间。我们的评估是基于控制器放置和迁移时间的详细模型。结果证实了迁移时间是一个关键参数,并且表明在只允许较短迁移时间的情况下,控制器的数量并不能带来更大的灵活性。对于宽松的迁移时间,更多的控制器可以带来更多的灵活性。结果还表明,动态适应流量变化,可以提供比非动态SDN控制平面高达60%的改进。
In Software Defined Networking (SDN), for performance reasons the control plane consists of multiple SDN controllers that provide a logically centralized network control. To react to traffic changes, the control plane may adapt to improve the performance of the whole network. This adaptation includes the migration of controllers, i.e. the change of their placement, as well as the change of switch to controller assignments. We call such adaptive network control a dynamic SDN control plane. Whereas most state of the art focuses on an optimal placement of controllers considering static traffic demands, we draw the attention to the need for a dynamic control plane and the performance of the adaption process itself. The involved controller migration and switch re-assignments significantly affect the control plane performance. Moreover, in this work we evaluate the flexibility of a dynamic SDN control plane with respect to the number of controllers. In particular, we refer to a flexibility metric that takes the time into account that is needed for successful migrations to handle traffic changes. Our evaluations are based on detailed models of controller placement and migration time. The results confirm that the migration time is a critical parameter and reveal that in case only a short migration time is allowed, the number of controllers does not bring more flexibility. For relaxed migration times, a larger number of controllers leads to more flexibility as expected. The results also show that a dynamic adaptation to traffic changes, can provide an improvement of up to 60% over a non-dynamic SDN control plane.