-Rob

“

The pre-beta version of Hyper-V 3.0 boasts a number of new and improved features, including a virtual machine replication technology known as Hyper-V Replica. It uses the Hyper-V Volume Shadow Copy Service (VSS) writer to prepare a snapshot of the virtual machine (VM) on the primary host (which can remain active) and moves the data to the secondary host. This capability adds disaster recovery capabilities to Hyper-V 3.0, allowing a virtualization administrator to have cold, ready copies of VMs anywhere in the world.

Hyper-V Replica creates copies on a VM-by-VM basis.While it is possible to replicate every VM to an alternate host with Hyper-V Replica,  in my experience, the tool is better suited to replicating one-off line-of-business VMs that provide important infrastructure fault tolerance, rather than full virtual environments. As such, Hyper-V Replica may be impractical for large enterprises with hundreds or thousands of VMs. In these situations, administrators can use PowerShell to configure multiple VMs at once.

“

Read the rest of the article HERE.

In the screen shot below I am viewing all VMs sorted by CPU usage.  As you can see there are some VMs that seem out of place.  XP_Test is listed at 0% in the GUI, but in the number 4 position of heaviest CPU usage.  It is also off which makes sense that it would be 0%, but not the 4th highest CPU usage.

This got me thinking about where all the data for SCVMM comes from, SQL.  I have edited SCVMM SQL tables many times to correct other issues so it makes sense that this is where I might be able to dig into why XP_Test is reporting as it is.

Within the SCVMM SQL tables performance data is not listed under its VM name, but instead is associated with its VMID.  In order to find the data I want, I need to first get the VMID.  The SCVMM PowerShell command below will get this VMID.

Once you have the VMID, it all comes down to knowing what table the SCVMM VM performance data lives (WLC_VMInstance), and using a SQL SELECT statement from within the SQL Management Console to find the information.

Looking at the results section from the SQL query, you can see that my XP_Test VM is reporting 31% CPU utilization, which, if that value was valid, would have put it 4th overall in my CPU Utilization.  Since the system is off, something went wrong with reporting correctly to SQL even though the GUI has the value correct at 0%, but incorrectly sorted due to the incorrect value in SQL.

By no means is this a large bug and the correction would be to manually modify table entry , but this is a good example of where the real data for SCVMM resides and why you should be knowledgeable about SQL if you use SCVMM.  Knowing how the tables are laid out and where specific data resides will help you tremendously in understanding the product and correcting occasional incorrect statuses or performance data.

“

For small IT shops, PowerShell cmdlets improve on Hyper-V Live Migration functionality. But, with the addition of System Center Virtual Machine Manager and its PowerShell capabilities, those shops can precisely coordinate live migrations between nodes and Cluster Shared Volumes.

Microsoft System Center Virtual Machine Manager (SCVMM) includes a robust graphical user interface (GUI) from which administrators can perform Hyper-V live migrations. Even so, you may feel that SCVMM lacks the granular control needed to administer a complex, Hyper-V virtual infrastructure.

SCVMM PowerShell cmdlets, however, provide increased flexibility when managing Live Migration, far beyond what is possible with the GUI. With the following SCVMM PowerShell cmdlets and scripts, for example, you can live migrate an entire host’s worth of virtual machines (VMs) to a specific node, or transfer VMs based on their Cluster Shared Volumes assignments.

”

For the rest of the article click HERE

It is good to see the number of Hyper-V application vendors starting to increase.  I hope we see more from Veeam and others soon.

-Rob

Running Cluster Validation returns these warnings.

(Click Picture for Larger View)

For the Quorum, the change is easy.  Just open the properties of the Quorum disk resource and on the Advanced Policies tab check the check mark next to the node to make it a Possible Owner.

(Click Picture for Larger View)

CSV disks this can be a bit more challenging however since you can not add or change the Possible Owner of a CSV from the GUI.  Luckily you can do this using the CLUSTER.EXE utility as outlined in Gustavo’s Post.

CLUSTER.EXE . RES “Cluster Disk 1″ /ADDOWNER:<nodename>

You could also use a PowerShell command to do the same thing.

• Open PowerShell
• Load the FailoverClusters Module
• Import-Modules FailoverClusters
• Run the Set-ClusterOwnerNode command:
• Get-ClusterResource "Cluster Disk 1" | Set-ClusterOwnerNode -Owners <nodename>

Looking at the Cluster Validation report again now will show that there are no problems with incomplete disk ownership.

(Click Picture for Larger View)

“

I am thankful for a Hyper-V 3.0 feature, but it’s not the flashier, increased number of virtual CPUs, the mammoth amounts of virtual machine (VM) memory, the concurrent Live Migration capabilities, or even the new VM replication feature.

Building on last year, I am thankful for the updated snapshot feature. The ability to roll back to a previous point in time after a botched application upgrade has always been a welcomed feature. But it came with a caveat: After you deleted a snapshot, you needed to shut down the VM to merge the AVHD file changes into the parent Virtual Hard Disk.

This requirement caused unwanted downtime for the application. Hyper-V 3.0 snapshots still provide the same valuable rollback method as before. But now the merging process happens live, without the necessary downtime. Thus, it eases the direct administrator interaction during application upgrades.

”

See the responses from the other Search Server Virtualization Advisory Board members here.

If you are trying to replicate a VM from one Hyper-V 3.0 host to another, you must make sure the the appropriate Network Resource Pools have been configured exactly the same on each node.  If your VM is part of a Network Resource Pool that does not exist on the replica server,  you will get the message below.

- Rob

Hyper-V Live Migration allows admins to move a  virtual machine (VM) from one Hyper-V cluster node to another with no noticeable downtime, but some shortcomings still remain, include the inability to live migrate multiple VMs at one time, the lack of scheduling capabilities for live migration, and the inability to migrate VMs based on specific criteria. Hyper-V 3.0, slated for some time in 2012, promises to improve on Hyper-V Live Migration’s limitations, but for now, there are viable workarounds with the help of PowerShell cmdlets.

Using the PowerShell Failover Cluster cmdlet
Most large Hyper-V clustered installations implement System Center Virtual Machine Manager (SCVMM), but for organizations with only one or a few Hyper-V clusters, PowerShell cmdlets can automate and orchestrate the live migration process, while avoiding the additional cost of SCVMM. Most Windows cluster administrators are familiar with Cluster.exe for managing cluster resources, but the PowerShell Failover Cluster cmdlet can accomplish even more live migration functions.

To read the rest of the article see the scripts go HERE.

After trying a more an more common task of modifying the host status within VMM database with SQL Management Studio without luck, I came across Chris’s Blog who had the solution to the same problem and I wanted to highlight his work. 

After a bit of VMM database surgery, the Host entry was cleaned up perfectly and all is now stable and working perfectly.  Thanks Chris.

Go HERE for the detail of this solution.

-Rob

GREAT BIG HYPER-V SURVEY of 2011

UPDATE:  Results are in.  Below is a link to the report and raw data from Aidan Finn’s site. 

http://www.aidanfinn.com/?p=11531

Rob