{"id":12729,"date":"2012-06-08T21:59:08","date_gmt":"2012-06-08T20:59:08","guid":{"rendered":"https:\/\/aidanfinn.com\/?p=12729"},"modified":"2012-06-08T21:59:08","modified_gmt":"2012-06-08T20:59:08","slug":"what-is-the-correct-number-of-simultaneous-live-migrations","status":"publish","type":"post","link":"https:\/\/aidanfinn.com\/?p=12729","title":{"rendered":"What Is The Correct Number Of Simultaneous Live Migrations?"},"content":{"rendered":"

Seeing as a VMware marketing employee was so kind as to link to this post from his “independent” blog, I’ll gladly inform his followers of his employers’ desperation<\/a> and that VMware storage is insecure<\/a>.\u00a0 You’re welcome, Eric \ud83d\ude42<\/em><\/p>\n

We have the ability to run many live migrations at once between Windows Server 2012 (WS2012) Hyper-V hosts.\u00a0 By default, we can do 2 simultaneous Live Migrations at once: that\u2019s 2 live migrations and 2 storage migrations.<\/p>\n

A word of warning.\u00a0 Storage live migrations cause a spike in IOPS so you don\u2019t want to do too many of those at once.\u00a0 Anyway, they are more of a planned move, e.g. relocation of workloads.\u00a0 The same applies to a shared-nothing live migration where the first 50% of the job is a storage migration.<\/p>\n

In this post, I\u2019m more interested in a traditional Live Migration where the storage isn\u2019t moving.\u00a0 It\u2019s either on a SAN or on an SMB 3.0 share.\u00a0 In W2008 R2, we could do 1 of these at a time.\u00a0 In vSphere 5.0, you can do 4 of these at a time.\u00a0 Windows Server 2012 hasn\u2019t applied a limit, just a default of 2.<\/p>\n

\"image\"<\/a><\/p>\n

So how many simultaneous live migrations is the right amount?\u00a0 How long is a piece of string?\u00a0 It really does depend.\u00a0 Remember that a VM live migration is a copy\/synchronisation of the VM\u2019s changing memory from one host to another over the network<\/em>, followed by a pause on hostA with an un-pause on hostB.\u00a0 Harder working VMs = more memory trash = longer process per VM.\u00a0 More bandwidth = the ability to run more migrations at once.<\/p>\n

So I decided to run a quick test tonight.\u00a0 This afternoon I build a converged fabric cluster using WS2012 RC.\u00a0 It\u2019s based on 4 * 10 GbE NICs in a team, with 9014 sized jumbo frames enabled, and I have a QoS poicy to guarantee 10% of bandwidth to Live Migration.\u00a0 I deployed 60 * 512 MB RAM Ubuntu VMs.\u00a0 And then I ran 4 sets of tests, 3 in each set, where I changed the number of concurrent live migrations.\u00a0 I started with 60 simultaneous, dropped to 20, then to 10, and then to the default of 2.<\/p>\n

\"image\"<\/a><\/p>\n

Here\u2019s a summary on the findings:<\/p>\n