{"id":8722,"date":"2015-08-13T09:57:36","date_gmt":"2015-08-13T07:57:36","guid":{"rendered":"http:\/\/blog.zhaw.ch\/icclab\/?p=8722"},"modified":"2015-08-13T10:32:44","modified_gmt":"2015-08-13T08:32:44","slug":"experimental-evaluation-of-post-copy-live-migration-in-openstack-using-10gbs-interfaces","status":"publish","type":"post","link":"https:\/\/blog.zhaw.ch\/icclab\/experimental-evaluation-of-post-copy-live-migration-in-openstack-using-10gbs-interfaces\/","title":{"rendered":"Experimental evaluation of post-copy live migration in OpenStack using 10Gb\/s interfaces"},"content":{"rendered":"

<\/b>Up to now, we have published several blog posts focusing on the live migration performance in our experimental <\/span>Openstack<\/span><\/a> deployment –\u00a0performance analysis of post-copy live migration in Openstack<\/a>\u00a0and\u00a0an analysis of the performance<\/a>\u00a0<\/span>of live migration in Openstack<\/span><\/a>. While we analyzed the live migration behaviour using different live migration algorithms (read our previous blog posts regarding pre-copy and post-copy (hybrid) live migration performance) we observed that both live migration algorithms can easily saturate our 1Gb\/s infrastructure and that is not fast enough, not for us!\u00a0Fortunately, our friends Robayet Nasim and Prof. Andreas Kassler from <\/span>Karlstad University<\/span><\/a>, Sweden also like their live migrations as fast and reliable as possible, so they kindly offered their 10 Gb\/s infrastructure for further performance analysis. Since this topic is very much in line with the objectives of the <\/span>COST ACROSS action<\/span><\/a> which both we (ICCLab!) and Karlstad are participants of, <\/span>\u00a0this analysis \u00a0was carried out under a 2-week short term scientific mission (<\/span>STSM<\/span><\/a>) within this action.
\n<\/span>This blog post presents a short wrap-up of the results obtained focusing on the evaluation of post-copy live migration in OpenStack using 10Gb\/s interfaces and comparing them with the performance of the 1Gb\/s setup. The full STSM report can be found here<\/a>.<\/span><\/p>\n

STSM Overview<\/b><\/p>\n

The STSM took place in Karlstad University in Sweden from 18.4.2015 to 2.5.2015 under supervision of Prof. Andreas Kassler. It was funded via the COST ACROSS (IC 1304) action. The main goal of the STSM was to evaluate state\u00ad of \u00adthe \u00adart virtual machine (VM) live migration approaches using the testbed available in Karlstad which has 10 Gb\/s network infrastructure. As a part of this STSM we also scheduled a short visit to the Distributed Systems lab led by Prof. Erik Elmroth at Umea University in Sweden as its researchers currently are one of the top contributors working on advanced virtual machine live migration techniques and they are also affiliated with the COST ACROSS action.<\/span><\/p>\n

Live migration basics<\/b><\/p>\n

First of all a short introduction \u00a0to the theory of live migration. Virtual machine (VM) live migration is a process of transferring a running VM from one hypervisor to another. Desirably, this process should be fast, convergent, and without any service interruptions. Currently, using the <\/span>QEMU<\/span><\/a> hypervisor you will probably deal with the built in pre-copy live migration algorithm that has a serious non-convergence risk which manifests when migrating VM with memory intensive loads. With a little more effort you can set up your environment to use post-copy (hybrid) live migration instead. As the post-copy algorithm fetches missing VM\u2019s RAM pages after the VM on the destination is activated there is ensured convergence for the \u00a0price of fragmentation of the VM\u2019s memory between the source and destination while migrating. The hybrid approach combines both solutions described to ensure convergence and reduce the fault sensitive post-copy phase. <\/span><\/p>\n

Setting up the environment and test framework<\/b><\/p>\n

That\u2019s the theory – now to the practice!. After my arrival and very nice welcoming in surprisingly sunny sweden we spent few days setting up the testbed.<\/span><\/p>\n

We used two physical machines <\/span>Dell PowerEdge M610<\/span><\/a> (Intel Xeon E5520 @ 2,7GHz, 24GB RAM) as a modules of a Blade server providing 10Gb\/s network interfaces through its backplane. We installed Openstack <\/span>Juno<\/span><\/a> with QEMU \/ <\/span>libvirt<\/span><\/a> with the post-copy support. One server was \u00a0configured purely as a compute node running nova and nova-network services. The second one was configured as both a compute node and the controller node providing also all the other management services.<\/span><\/p>\n

Evaluating the live migration performance we focused on measuring:<\/p>\n