Storage Design

Using Differencing Disk to Reduce Storage Cost

A differencing disk is a virtual hard disk used to isolate change to a virtual hard disk or the guest operating system by storing them in a separate file.Using the technology of differencing disk, all the VMs references to the same parent VHD and only saving changes in child VHD, so the operation in one VM won’t affect others. List the following arthitecture of using Differencing disk in Training Cloud solution:

Install Failover Cluster using MS iSCSI Target

MS iSCSI Target Server is integrated into Windows Server 2012 Berta as a role of File and Storage Services.MS iSCSI Target server is used to provide storage for remoting server, no matter it is a physical or virtual one.Through iSCSI initator service, remoting server regards VHDs on iSCSI Target server as its local storage, in the high speed LAN of Training Cloud envrioment.

File Share Storage over SMB 2.0

Hyper-V of Windows Server 2012 Beta supports use of SMB2 file shares to provide storage for virtual machines. This new feature supports a new way to provide virtual machines with shared storage, without use of a storage area network(SAN), this technology is one application of Network Area Storage(NAS). The following figure shows SMB2.0 File Share Storage used in Training Cloud solution:

4.4.4. Live Migration on Windows Server 2012 Beta

Live Migration allows a server administrator to move a running vitual machine or application between different physical machines without down time. As a successful live migration, the memory, storage, and network connectivity of the virtual machine needs to be migrated to the destination.
Live Migration has been improved greatly in Windows Server 2012 Beta. In past,live migration is only supported of high available VMs between cluster nodes. But now, scenarios of live migration are increased widespreadly, totally four types of live migration are supported in new Hyper-V:
  • Live Migration Without Shared Storage
  • Live Migration with SMB Shared Storage
  • Live Migration with Failover Cluster
  • Live Storage Migration
In this solution, lab VMs are stored on SMB shared storage as high available.If physical server shut down caused by unexpected failure, however it is inevitable in datacenter, the lab VMs will live migrate automatically to other suitable nodes. Users will experience the continuous service since the whole migration procedure. After discharging the failure, administrator then migrate the VMs back to first physical server with zero downtime.

Last edited May 25, 2012 at 2:20 AM by xinwenting, version 4


No comments yet.