Server 2012 R2 Hyper-V Replication and Deduplication

March 18, 2016, 2:19 am

≫ Next: How to completely remove replication on target?

≪ Previous: VM Console Shows "CTRL+ALT+DEL" Instead of "Installing Updates" Progress

$

0

0

Morning All

I am currently setting up a backup server which will have a replica copy of our main Hyper-V VM's replicated automatically using the Hyper-V replication feature. (Every 5 mins)

All VM's are running Win 2012 R2 Server, some of these are file servers, other IT servers for AV etc.

I'm considering enabling Server 2012 R2 deduplication on the volume that the VHDx files are stored on, on the Replica server. I think quite a few of these VHDx files will contain easily deduplicated data, saving me quite a lot of space.

As this is only a backup server, should there be any issues with this?

Deduplication is not enabled inside the VM's, I had too many issues trying to backup the VM servers with deduplication enabled, it hammered the drives too much.

Thanks in advance for your thoughts and suggestions. :)

---

How to completely remove replication on target?

March 1, 2016, 2:28 am

≫ Next: error 0x8000FFF in hyper-v windows 2008 server R2

≪ Previous: Server 2012 R2 Hyper-V Replication and Deduplication

$

0

0

Hyper-V replication sometimes causes problems. What I'm usually doing in this case is to remove and re-enable replication for a VM, and that solves the problem.

Have customers with Hyper-V Clusters. VM replicas are performed to the replica broker role on a target clusters. When a VM has replication errors, I remove replication from the VM and remove the replicated VM on the target cluster. But since some time, re-enabling replication fails with errors "file in use" or "vm is not in the current state to accept replication". 

Despite we have disabled replication and deleted the replicated VM on the target cluster, the XML file from the replicated VM is still in c:\clusterstorage\volume1\Hyper-V Replica\virtual machines\<vm-guid>.xml. It's not possible to delete this file, even when stopping the "Hyper-V Virtual Machine Management" Service on the cluster node that owns the CSV. The only "drastic solution" that works is to put the CSV into maintenance mode, add then a drive letter to the disk, and then delete the <vm-guid>.xml file. Then, after taking the CSV out of maintenance mode, replication of the problem VM can be re-enabled successfully.

What is the right way to completely remove a VM replication? How can this problem beeing solved without taking the CSV offline? Thank you in advance for any help.

Franz

error 0x8000FFF in hyper-v windows 2008 server R2

March 15, 2016, 7:03 am

≫ Next: Standalone Hyper-V Configuration

≪ Previous: How to completely remove replication on target?

$

0

0

Hello,

here is my problem: when I start my VM, I receive the error "fail to restore" with ID 0x8000FFFF.

I tried to apply the fix described in this page: https://support.microsoft.com/en-us/kb/2960802

However, the command "Net start vmhostagent" returns the error "invalid service name".

Is it because it only works on a windows server 2012 R2?

If I execute the command "Net stop vmms", it works, but my VM still sends the error 0x8000FFFF at start.

Do you have any idea of what I could do?

Thanks!

Standalone Hyper-V Configuration

March 18, 2016, 3:26 am

≫ Next: 2 node cluster: 2008 R2 to 2012 R2 hosts

≪ Previous: error 0x8000FFF in hyper-v windows 2008 server R2

$

0

0

Hi Team,

           I have  a New Dell Server with 04 Nic Cards. I would like to configure Hyper-V on the same, Server will be standalone.

         Kindly help on the configuring Hyper-V Network (Eg: Separate Network for VM, Management and Teaming etc..)

2 node cluster: 2008 R2 to 2012 R2 hosts

March 18, 2016, 1:11 am

≫ Next: hyper-v issue: Mouse pointer invisible in Linux Mint 17 (run in hyper-v of windows 10)

≪ Previous: Standalone Hyper-V Configuration

$

0

0

Hey,

We currently have a 2 node 2008 R2 cluster and we've got new hardware to replace these.  The new hardware is running 2012 R2.

What is the best process to migrate things over?  I actually thought I could add the 2012 R2 servers to the existing cluster and just live migrate VMs to them, then remove the old servers from the cluster.

No downtime then.  Nothing else is new (storage, networking etc.).

But on the Add Node wizard it says nodes must be running the same hardware, software, drivers etc. etc.  Can 2012 R2 servers be added as nodes to a 2008 R2 cluster?

Thanks

hyper-v issue: Mouse pointer invisible in Linux Mint 17 (run in hyper-v of windows 10)

November 4, 2015, 4:08 am

≫ Next: New VM, Gen1 or Gen2...?

≪ Previous: 2 node cluster: 2008 R2 to 2012 R2 hosts

$

0

0

Hi everyone:

    my problem is, after I installed the Linux mint 17 in windows 10 hyper-v, the mouse was invisible in the guest OS (Linux Mint 17). While the mouse really works, I can click the buttons in the guest OS, it's just invisible. anyone know what happened? I've tried to download the Hyper-v Linux integration services and tried to install it in Linux Mint and failed to do it. The setting of hyper-v shows that all the "integration services" are offered except the "guest services".

New VM, Gen1 or Gen2...?

March 18, 2016, 10:28 am

≫ Next: Hyper-V Replica Broker stuck in "Failed" status.

≪ Previous: hyper-v issue: Mouse pointer invisible in Linux Mint 17 (run in hyper-v of windows 10)

$

0

0

HI,

Years ago, I had VM running Exchange 2010 running on Windows 2012 on a HyperV 2012 host.  I upgraded the host to 2012R2, no problems.

6 months ago, I rebuilt the VM mail server from scratch on a Windows 2012R2.   Now noticed that the vhdx file is booting off the IDE controller.  I assume this is a Gen 1 VM.

Now I want to move this VM to a newly create Win2012R2 server.

My question is... when I create the new VM for that vhdx file, can I create a Gen 2 vs a Gen 1 vm...?  After created, I will copy the vhdx file to the host server and add the disk as I normally would...?

Or do I have to stick with the Gen 1 as the vhdx looks like that is how it was created...?

Thanks... Tom...

---

Tom Karpowski...

Hyper-V Replica Broker stuck in "Failed" status.

May 1, 2013, 2:55 pm

≫ Next: No full screen in Windows XP console running as Hyper-V guest OS.

≪ Previous: New VM, Gen1 or Gen2...?

$

0

0

2 2012 standard servers in a cluster. Trying to add the replica broker so I can have my HA VMs replicated off site as a DR solution. I've tried to install the role over a dozen times, with a dozen different names, no nice.

I've changed the OU security to include the cluster's computer having the rights to create computer accounts. I've even tried to "stage" a computer before adding the role but it just complains that it is already in use in active directory. I'm at the end of my rope as to why this isn't working.

 

I've done everything that has been suggested in these forums and the blogs I could find on google.

 

Any help would be great. Thanks guys!

---

No full screen in Windows XP console running as Hyper-V guest OS.

March 19, 2016, 12:40 pm

≫ Next: Virtual machine on hyperv 2012r2 backup problem

≪ Previous: Hyper-V Replica Broker stuck in "Failed" status.

$

0

0

I have an old DOS program that runs fine in a console window under native Windows XP. I've setup a shortcut on the XP machines and selected full screen. When the application runs it opens full screen.

On a Windows 10 with a Hyper-V instance of Windows XP the application is working just fine however no matter what I've tried it runs in a too-small box. The settings on the shortcut are the same as the ones on the native XP boxes that work.

Any Ideas?

Virtual machine on hyperv 2012r2 backup problem

March 19, 2016, 10:05 am

≫ Next: No full screen in Windows XP console running as Hyper-V guest OS.

≪ Previous: No full screen in Windows XP console running as Hyper-V guest OS.

$

0

0

Hi there.

We have Hyper-V on 2012 r2 HOST.

Virtual machine is also on 2012 r2 OS with Exchange EDGE 2013 role.
When we wish to make backup with VEEAM BACKUP AND REPLICATION (version 8 + update 3) we recieve an error.

3/19/2016 12:08:13 PM :: Unable to create snapshot (Microsoft Software Shadow Copy provider 1.0) (mode: Hyper-V child partition snapshot). Details: Writer 'Microsoft Hyper-V VSS Writer' is failed at 'VSS_WS_FAILED_AT_PREPARE_SNAPSHOT'.
The writer experienced a transient error.  If the backup process is retried,
the error may not reoccur.
--tr:Failed to verify writers state.
--tr:Failed to perform pre-backup tasks.

3/19/2016 12:08:34 PM :: Make sure VM does not have 'iSCSI Software Target Storage Provider' feature installed.

3/19/2016 12:08:34 PM :: Retrying snapshot creation attempt (Writer 'Microsoft Hyper-V VSS Writer' is failed at 'VSS_WS_FAILED_AT_PREPARE_SNAPSHOT'.
The writer experienced a transient error.  If the backup process is retried,
the error may not reoccur.
--tr:Failed to verify writers state.
--tr:Failed to perform pre-backup tasks.)

Any hints how to solve this guys?

With best regards

---

bostjanc

No full screen in Windows XP console running as Hyper-V guest OS.

March 19, 2016, 9:42 am

≫ Next: Direct Connected Storage versus VHDX File

≪ Previous: Virtual machine on hyperv 2012r2 backup problem

$

0

0

I have an old DOS program that runs fine in a console window under native Windows XP. I've setup a shortcut on the XP machines and selected full screen. When the application runs it opens full screen.

Edit: "I have the XP instance in the HYPER-V in fullscreen mode already. That Windows XP desktop fills the entire monitor. However, when I open a console window I can not get that window to use the whole screen like it does on the stand alone XP boxes."

On a Windows 10 box with a Hyper-V instance of Windows XP the application is working just fine however no matter what I've tried it runs in a too-small box. The settings on the shortcut are the same as the ones on the native XP boxes that work.

Any Ideas?

Direct Connected Storage versus VHDX File

March 19, 2016, 6:11 pm

≫ Next: Bridge external virtual switch to private virtual switch for performance

≪ Previous: No full screen in Windows XP console running as Hyper-V guest OS.

$

0

0

I'm building a file server on hyperv. The virtual machine itself as well as the individual files hosted on the file server will both be backed up to Azure. I have about 4Tb of files to migrate to this server. I'm having a hard time finding good info on best practices around the disks. Would it be better to just pass the physical disks into the VM or to use a 5TB VHDX file for the storage?

---

John Marcum | Microsoft MVP - Enterprise Client Management
My blog: System Center Admin | Twitter:@SCCM_Marcum | Linkedin: John Marcum

Bridge external virtual switch to private virtual switch for performance

March 19, 2016, 8:20 pm

≫ Next: Shrinking VHDX of Win 7 VM installed on Win Server 2012 R2

≪ Previous: Direct Connected Storage versus VHDX File

$

0

0

Scenario:

I have a large number of hyper-v guests (linux and windows) which exchange traffic amongst each other.  Additionally, on the same subnet, are several external servers which the hyper-v guests occassionally need to access.  Currently, all hyper-v guests connect to the physical adapter (via the standard external virtual switch) at 1gb/s.  Pretty generic stuff.

Question:

Is there any performance gain to be seen by creating a private virtual switch for all the hyper-v guests and implementing a transparent bridge between the private and external switches?  This would seem to allow me to decouple all the internal guests from the physical NIC and allow them to communicate between themselves at the full 10gb/s rather than being limited to the NIC's 1gb/s while preserving the ability to access external servers.

Shrinking VHDX of Win 7 VM installed on Win Server 2012 R2

February 29, 2016, 6:57 pm

≫ Next: Virtual SpinRite machine(s)

≪ Previous: Bridge external virtual switch to private virtual switch for performance

$

0

0

I want to reduce the size of a VM hard drive.  I’ve successfully converted it from VHD to VHDX,restarted the VM and ensured the vhdx is active and showing in the Hyper-V settings as the file & location of the VM.    

 

There are no snapshops or shadow copies. I've defragged the disk which shows 300 gig of the 450 gig is free and 0% fragmentation.   I’ve turned off the VM.  From the Win 2012 R2 Hyper-V host, when I open edit the settings of the vhdx disk file, I don’t get the option to shrink the file.  Only options are Compact, Convert or Expand.   Is it because the drive to be shrink has the system on it? 

Thanks for any help

Virtual SpinRite machine(s)

September 25, 2011, 10:15 am

≫ Next: Hyper-V applying snapshot after VM was removed

≪ Previous: Shrinking VHDX of Win 7 VM installed on Win Server 2012 R2

$

0

0

I am trying to set up a computer to handle multiple VM's, each running SpinRite on a single drive attached to the mother board.

I have that part all working, in the sense that SpinRite starts up and detects the physical disk I've passed through to it from Server 2008 R2. The issue is that SpinRite believes the VM is reporting the disks size incorrectly so it won't run. Do VM's do things to the reporting of passed through disks that could account for this?

---

Hyper-V applying snapshot after VM was removed

March 20, 2016, 4:40 am

≫ Next: Hyper-V Edit Disk with Checkpoints

≪ Previous: Virtual SpinRite machine(s)

$

0

0

Hi all,

I've found that strange hyper-v host server behavior.

After playing with some guest VM, I've decided to delete it and start with a new one.

I've removed it from Hyper-v management consoler, and tryed to delete vhdx files from its folders, but got an error that those files are in use.

I've checked resmon, and found that System process applying snapshots from deleted VM!

Is that a bug, or that's supposed to be like that?

host VM server is 2012R2, guest VM was also 2012R2.

Hyper-V Edit Disk with Checkpoints

March 20, 2016, 11:35 am

≫ Next: Remote Hyper-V Manager: Unable to establish communication between server and client

≪ Previous: Hyper-V applying snapshot after VM was removed

$

0

0

I needed to increase the C drive on a VM. I went to Edit Disk and changed the size. (I know.. now).

I found several references online to a tool called VHDTool.exe and found several copies but none of them included the /repair switch that was needed. 

I deleted the checkpoint and the VM C drive to point to C.vhd instead of the C_xxxxxx.vhd and the machine. The machine booted and the files for the E.vhd started "merging". 

Everything appears to be OK except I noticed on my network monitoring software the workstations were having troubles with WMI connections and then I noticed one of the workstations having domain trust relationship errors. 

Right now I have C.vhd, C-xxxx.vhd, E.vhd. 

What are your thoughts?

Remote Hyper-V Manager: Unable to establish communication between server and client

February 26, 2014, 2:41 am

≫ Next: VM Resource Metering

≪ Previous: Hyper-V Edit Disk with Checkpoints

$

0

0

Hi everybody,

I know that my problem is not a new one and there are tons of post and blogs that talking on it. But it seem that my case is a little different because all solutions that I found doesn't work.

Let present the case: I have 2 Hyper-V Server 2012 in a datacenter (call it Host1 and Host2). The datacenter is connected to my office though a VPN managed by our ISP. From my computer in my office (Call it Test and he run under Windows Server 2012), I can remotely manage both servers with Server Manager (File, Storage, Hyper-V, ...). When I open Hyper-V Manager, I can only manage the VM running on Host1. For Host2, I receive the message "RPC Server Unavailable. Unable to establish communication between"Host2" and "Test". But I can manage Hyper-V host settings, Virtual Switches, ... The only things that I can manage is the VM running on this host.

The problem is not present between the 2 host.

After reading a lot of things, I not find how I can resolve this. Somebody can help me?

Thanks in advance.

Regards,

Jerome (from Belgium)

VM Resource Metering

March 21, 2016, 12:30 am

≫ Next: Replication over 10Gbps links only using 1Gbps

≪ Previous: Remote Hyper-V Manager: Unable to establish communication between server and client

$

0

0

I have resource metering enabled on my VMs on a 4-node Hyper-V cluster. I have not created any additional pools, so they are all in primordial. I have processor resource monitoring enabled against this pool.

Last week I was using: 

Measure-VMResourcePool -ComputerName $Hosts -Name Primordial -ResourcePoolType Processor 

It was returning very low values. I ran: 

get-vm -ComputerName $Hosts | Measure-VM | Select-Object VMname,avgcpu 

This showed a lot of the VMs with a null entry for the AvgCPU.

I ran (again): 

Enable-VMResourceMetering -ResourcePoolName Primordial -ResourcePoolType Processor 

Then the measure-vm command again. This time it showed AvgCpu for all my VMs.

Following patching and reboot of the hosts over the weekend, I have seen similar behaviour where most VMs were showing null values for AvgCPU. Running the Enable-VMResourceMetering command again has resolved it.

Is this expected behaviour or do I have an issue? It seems that maybe new VMs do not start metering their CPU, even if it's already enabled on the pool; and also that the setting is lost on most VMs after reboot.

If it is expected, will it work any differently if I create my own resource pool?

I have a workaround for the new VM issue and I can easily put in a startup script for the host reboot issue, but I would like to know if it's a real issue or just the way it works.

Replication over 10Gbps links only using 1Gbps

March 21, 2016, 6:34 am

≫ Next: Hyper-V error l

≪ Previous: VM Resource Metering

$

0

0

Hey everyone,

I'm trying to wrap my head around something and determine if there is an issue with my server, the network, or if things are working as designed.

Here is my setup:

I have 2 Hyper-V servers; 1 primary and a second one is the replica. I have 5 VMs (6.2TB storage per VM) that live on the primary and I need to have them replicate to the replica server. I have 3 sets of teamed NIC in Switch Indepentent mode acting as active/passive. One team (2 x 1Gb NICs) is for hypervisor management. One team (2 x 10Gb NICs) are for the VM traffic, and one team (2 x 10Gb NICs) for replication traffic. On the physical network side, the traffic goes through 2 10Gb fiber switches.

When I start the VM replication over the network, I can see the traffic go through the Replication team, but the data sent stays between 800Mbps and 1.2 Gbps. Is this speed normal for replication, normal for teaming NICs, or is there something I am missing?

Thanks,

Josh