Live Migration of Virtualized Edge Networks: Analytical Modeling and Performance Evaluation

Following the current evolution of virtualization techniques and software defined networking, edge networks might evolve towards a fully virtualized implementation by means of a number of virtual machines working cooperatively to perform the tasks of existing network middleboxes. In such a scenario the possibility to migrate groups of cooperating virtual machines as a whole set may be a very important feature, but what will be the performance issues of this solution? The live migration performance of multiple virtual machines working in some sort of correlated manner is a topic that has not been widely studied in the literature. This manuscript presents a model reasonably simple to implement that may be used to derive some performance indicators such as the whole service downtime and the total migration time. The model is used to compare some scheduling strategies for the migration and provide guidelines to such implementation.

[1]  Vyas Sekar,et al.  Making middleboxes someone else's problem: network processing as a cloud service , 2012, SIGCOMM '12.

[2]  Gunjan Tank,et al.  Software-Defined Networking-The New Norm for Networks , 2012 .

[3]  Andrew Warfield,et al.  Live migration of virtual machines , 2005, NSDI.

[4]  Khaled Z. Ibrahim,et al.  Optimized pre-copy live migration for memory intensive applications , 2011, 2011 International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[5]  Franco Callegati,et al.  Clouds of virtual machines in edge networks , 2013, IEEE Communications Magazine.

[6]  Hai Jin,et al.  Performance and energy modeling for live migration of virtual machines , 2011, Cluster Computing.

[7]  Mark Handley,et al.  Is it still possible to extend TCP? , 2011, IMC '11.