Part II: VMware vMotion vSphere 6.0

Do check PART- I and PART-III in this series.

What is VMware vMotion:

VMware VMotion feature enables the live migration of running virtual machines from one physical server to another with zero downtime and continuous service availability.


Image: VMware

VMware vMotion allows users to perform hardware maintenance without downtime. Proactively migrate virtual machines away from failing host or from heavily overloaded servers to balance the workload.

When we talk about vMotion migration, it does not mean that we are migrating complete VM including VM files. Here we are talking about moving only the CPU and Memory content from one ESXi host to another. VM files on shared storage are untouched except that the ESXi ownership is changed.

Also, we need to understand few points before we explore vMotion feature.

When we talk about server downtime’s, there are two terms into it which we typically use, namely planned downtime and unplanned downtime. Lets see where vMotion fits in this chart.

As you can see in above chart, vMotion in itself is a manual process which means administrator has to initiate vMotion manually to migrate the virtual machine from one server to another. However, it can be automated by using DRS functionality in fully automated mode.

vMotion does not provide high availability in unplanned downtime, we have to use other means of high availability features like vSphere HA, Fault Tolerance (FT).

vMotion can be used to avoid downtime in planned downtime situations.

How Does VMware VMotion Work step by step? 


  • First of all it verifies the target server, if it can run the VM after it is migrated.
  • Once verified the target ESXi, it starts VM process on target ESXi and reserves the resources.
  • In next step, system memory checkpoint is created for source VM.
  • In next step previously written contents of the system memory checkpoint are transferred to the target VM also referred as Pre-copy. 
  • Now during previous steps, it takes time to copy memory content though its small, so there is possibility that some new memory pages are created on Source VM called as Dirty pages. 
  • To copy these new pages to target VM, the source VM is stunned fort short while, i.e. CPU of the source VM is stopped. 
  • The dirty pages (bitmap) of source VM memory are transferred to the target VM. 
  • The vMotion process is completed and a Reverse ARP packet is sent back to confirm the transfer process completion. 
  • VM files ownership on shared storage is taken by the new ESXi host. 
  • The source VM deleted on first ESXi host.


What is new in vSphere 6.0 vMotion?

 
Cross vSwitch vMotion

It allows you to migrate virtual machines between different vSwitches. Below is the list of the same.

    • Standard switch to Standard switch
    • Standard switch to Distributed vSwitch
    • Distributed vSwitch to Distributed vSwitch (same version)


Below two vMotion’s are not supported.

    • Distributed vSwitch to Distributed vSwitch (different version)
    • Distributed vSwitch to Standard vSwitch
  • Cross vCenter vMotion

We can move virtual machine between vCenter Servers. This is true not only for vMotion but you can do this with all the different migration types like storage vMotion, network vMotion. In fact this is possible without having a shared datastore between the source and destination vCenter. This concept is known as shared nothing migration.

For this type of migration, source and destination vCenter Server must belong to the same SSO domain.
 
Long Distance vMotionLong distance vMotion uses socket buffer resizing techniques to ensure that migrations succeed when latency is high. The maximum supported RTT latency is 150ms.

  • vMotion Network improvements:

L2 adjacency for your vMotion network is no longer required as vMotion traffic is now fully supported over an L3 connection.

  • vMotion support for Microsoft Clusters using physical RDMs. 

This is supported with Windows 2008, 2008 R2, 2012 and 2012 R2.

Leave a Reply