Hi guys, I'm doing a new hyper-v setup, and have an external HPE MSA 2052 SAN storage. When provisioning storage for the VMs,  which is the recommended way to host my VMs (VHDs and Config files), in a single large LUN/volume, or create separate LUNs/volumes for each VM?

Is  there any disadvantages of either options?

I got a question about SSD performance of Hyper-V clients.

I used an Intel 900P 280GB drive as the storage for two Hyper-V clients, both running Windows 10 Pro 1809. If I only power one of them on, and I use CrystalDiskMark 6 to test the disk performance, the Q1T1@4K (Queue 1 Thread 1@4K) result (the one that made all the difference with Optane) will be only Read:110MB/S Write:105MB/S, but if I test the same drive in the host system, the result will be Read:235MB/S Write:225MB/S.

I tried different platform, and the result is identical, the disk performance of clients are seemingly limited at 110MB/S.

I managed to set up the entire 280GB drive to direct pass through as a SCSI drive for one client, and the disk performance is still like above, halved the test result within host system.

The sequential read and write performance is the same for clients and the host, but that does not make the biggest impact on the smoothness of the system. I wonder if there are some tweaks to change this.

My host system configuration is: AMD Ryzen 2600, 8-core, 64GB DDR4 2666Mhz, Intel 900P 280G NVME

Hello dear Hyper-V elites. I'm in need of help. I've googled without success, I can't find anything matching this issue.

We have a customer with 2x Hyper-V 2016 machines. HOST-A and HOST-B. In this case the issue seems to originate from HOST-B. 5 of the servers that is currently running on HOST-B and is set to replicate to HOST-A is giving us headaches.

The 5 servers are all running Linux (Ubuntu), and we have no issues with the other VM's that are running Windows. What we receive are alerts with the following description:

EventID: 19060

'Linux-Server-1' failed to perform the 'Cleaning up stale reference point(s)' operation. The virtual machine is currently performing the following operation: 'Backing up'. (Virtual machine ID XXXXXXX-XXXXXXX-XXXXXXX-XXXXXX1)

Directly after the following event-log entries comes up;

EventID: 19070 and after that 19080

'Linux-Server-1' background disk merge has been started. (Virtual machine ID XXXXXXX-XXXXXXX-XXXXXXX-XXXXXX1)

'Linux-Server-1' background disk merge has been finished successfully. (Virtual machine ID XXXXXXX-XXXXXXX-XXXXXXX-XXXXXX1)

However we never get the code 29174 which we receive from all other "working" windows VM's.

Example below;

Replica statistics for 'Windows-server-1' (Virtual machine ID XXXXXXX-XXXXXXX-XXXXXXX-XXXXXX2)
StartTime: 132184512000256627
EndTime: 132184943997080231
ReplicationSuccessCount: 144
ReplicationSize (bytes): 1194310144
ReplicationLatency (seconds): 3
ReplicationMissCount: 0
ReplicationFailureCount: 0
NetworkFailureCount: 0
MaxReplicationLatency: 1
Application-consistentCheckpointErrorCount: 0
PendingReplicationSize (bytes): 0
MaxReplicationSize (bytes): 79151616

They all run on the same backup solution (solarwinds msp backup & replication) and at the same schedule. We receive no errors on the backup console.

All servers are also "hosted/located" on the same disk. Since I've seen other topics mentioning that it could be a disk issue, we think this is not the case.

Help me obi-wan kenobi, you're my only hope.

//Regards, Andreas

We have several Windows 2012R2 hyper-v hosts with identical hw-config and all storage is local attached, virtualized on storage controllers over Fibrechannel. In addition all hosts has dual 10GbE converged network adapters. The hosts are domain members and no failover clustering is enabled.

What is the most effective way to move VM's between hosts in a SAN environment ? 
The "old" way of export/import requires significant downtime on large volumes.
The same goes for Live migration without SMB storage ?
Replication is much the same, i.e moving data over the fabric network ?

I've always used the features of the storage controller when moving vm's between hosts by shutting down, changed zoning and remapped the LUN's to the new host, created the VM's manually and attached existing disks - almost without downtime and no physical duplication on storage. The backup is done on the storage controller with flash or volume copy. 

Does this manual approach represent any risk on identical host hardware ?
The VM Bios GUID is of course changed, but I've never got any problems with reactivating licenses etc.
Btw I've never used ROK/OEM software on the VM's and both SQL Server and ADS/RDS was working flawless after moving.

But now I have to move a Exchange 2013 DAG member to a new host and have some doubts after reading several articles on this topic... The plan is as follows:

- Enable maintenance mode and stop/change startup type to disabled for all Exchange services, shutdown.
- Remap storage to new host and create VM in Hyper-V manager with identical settings
- Start VM and check network connectivity and event log
- If all OK, reset startup type on Exchange service and reboot
- Check events - disable maintenance mode

Is this a viable plan ? Any comments will be highly appriciated !

Cheers,

Terje

I have a client that filed to tell me their physical domain controller had two drives fail. The two domain controllers on their Hyper-V Core system continued to run, so they thought it wasn't important. Anyway, the VMs decided to do updates and both rebooted last night, possibly the Hyper-V server rebooted as well. Anyway, when they go in this morning no one could authenticate, get DHCP leases or access either the Internet or their intranet. It doesn't look like I'll be able to get their physical DC up any time soon, but the Hyper-V Core host is running fine. None of the four VMs on that server will start though, they all respond to a Start-VM with a "Failed to Initialize" error. When I request a dump from the event handler I see this, duplicated for each VM:

Get-WinEvent -FilterHashTable @{LogName ="Microsoft-Windows-Hyper-V*"; StartTime = (Get-Date)
.AddMinutes(-10); }


   ProviderName: Microsoft-Windows-Hyper-V-Hypervisor

TimeCreated                     Id LevelDisplayName Message
-----------                     -- ---------------- -------
11/19/2019 4:17:01 PM        16642 Information      Hyper-V successfully deleted a partition (partition 6).
11/19/2019 4:17:01 PM        16641 Information      Hyper-V successfully created a new partition (partition 6).


   ProviderName: Microsoft-Windows-Hyper-V-VMMS

TimeCreated                     Id LevelDisplayName Message
-----------                     -- ---------------- -------
11/19/2019 4:16:55 PM        15120 Error            'SRV-SHU' failed to initialize. (Virtual machine ID 723248F1-6C3...
11/19/2019 4:16:55 PM        14070 Error            Virtual machine 'SRV-SHU' (ID=723248F1-6C33-4BA8-A960-83CE0F157D...

I got a question about SSD performance of Hyper-V clients.

I used an Intel 900P 280GB drive as the storage for two Hyper-V clients, both running Windows 10 Enterprise 1809. If I only power one of them on, and I use CrystalDiskMark 6 to test the disk performance, the Q1T1@4K (Queue 1 Thread 1@4K) result (the one that made all the difference with Optane) will be only Read:110MB/S Write:105MB/S, but if I test the same drive in the host system, the result will be Read:235MB/S Write:225MB/S.

I tried different platform, and the result is identical, the disk performance of clients are seemingly limited at 110MB/S.

I managed to set up the entire 280GB drive to direct pass through as a SCSI drive for one client, and the disk performance is still like above, halved the test result within host system.

The sequential read and write performance is the same for clients and the host, but that does not make the biggest impact on the smoothness of the system. I wonder if there are some tweaks to change this.

My host system configuration is: AMD Ryzen 2600, 8-core, 64GB DDR4 2666Mhz, Intel 900P 280G NVME

I have 19 differencing VMs running in my Windows Server 2008 R2 host which all connect to a single Parent VM.

I want to migrate the 20 VMs to a Windows Server 2016 host.

I have read I can migrate the 19 differencing VMs and the parent VM but this may not be straight forward and a possible easier option would be to merge the 19 differencing VMs and the parent to leave 19 normal VMs.

The question is which option is best and what are the steps involved to perform thus chosen option.

I have a new installation of Server 2019 STD with Hyper-V role installed.  I am getting the following errors over and over:

"Failed to move RSS queue 1 from VMQ 3 of switch DCC49724-C426-4DF1-B10F-9533638F2C18 (Friendly Name: HPE Ethernet 10Gb 562SFP+ Adapter - Virtual Switch), ndisStatus = -1073741811 ."

The Guest VM is losing network connectivity every once in a while for a minute or so.

Is this error the cause, and if so how to resolve?

Brian Modlin

Hi All

Here is the setup. Both Virtual machines are setup for dynamic, in VM2 only shows 14GB MAX RAM. Why is that? Is it because that is how much RAM is using real time? 

HPV Host 64GB

 VM1: shows assigned memory at 8GB

 VM2: shows assigned memory at 14GB  Dynamic Max RAM is 1048576 MB

Thank you in Advance. 

Hi all!

Need your help for upgrading from Microsoft Hyper-V  2012 R2 to Microsoft Hyper-V 2016/2019;

During installation, the following error occurs:

The installation failed in the SAFE_OS phase with an error during SET_PRODUCT_KEY operation.

Kindly advice.

Thanks

I am able to use vTPM if it is enabled before OS install.  However for the existing Gen2 VM I get an error starting the VM after vTPM is enabled.

Sorry for the chopped up error text, I did not copy the entire message.

"RemoteException: failed to start worker process........the computed authentication tag did not match the input authentication tag"

I am at a loss here, not finding much going pages deep into Google searches on various combinations of the terms.  Anyone run into this or have any ideas on where I should go?

Have one machine in my VM environment that will not Live-Move due to this listed reason: "The virtual machine '*****' is using processor-specific features not supported on physical computer '******'. "  This is a SQL VM, and the builder did do some NUMA tuning on this VM.  It is running fine on the host it's on, and all other hosts in the environment are 100% identical? We are using failover cluster manager..

Thanks,

Jason

Greetings,

I have a Windows 10-  Pro Hyper-V virtual machine that I would like to mount on a Windows Sever 2016 Hyper-V host.

Is this possible?

Thanks

To make a long story short I have a 32 bit Win 10 pro VM which I am using to run a bunch of old but VERY VALID educational programs that are CD/DVD based. These programs were originally designed to work in Win 98 or XP all 32 bit.

I have created .ISO versions of the CDs and I mount them and run the old programs in a VM.

The way I want to have this work, and the way it DID WORK under the Win 7 "XP Mode" feature is as follows...

A user does a normal log in to the system, was Win 7 ENT, now Win 10 Ent, X64. On the 2nd screen was an Icon which win clicked would start the XP Mode VM, open it up in full screen and log in to the "guest OS" which only had Icons to run the learning software, which was the only software, other than basic "Windows stuff" and the software to mount the .ISOs of the learning programs installed . The user then would click on one of the Icons, the program would run, they would use it then close it and either start a new program or close the VM out. At no time would they have to log in to the VM.

The old XP Mode from Win 7 was an excellent tool for doing this and enabled the college I work at to continue to use the many perfectly valid learning/teaching/studying programs that they have with out needing to buy in to the new "pay for the rest of your life" greed based subscription based/ cloud based learning programs these products have mostly morphed in to.

If Win 10 had a similar XP Mode that would pass sound to the hosting system and could autologin like the XP Mode does on my Win 7 system I would not need to be here asking this question, but alas MS crippled the XP Mode, yes I can run the original XP Mode VM in Hyper-V and get the old programs running, EXCEPT for sound/resolution which MS has INTENTIONALLY disabled.

Fast forward....

I have a Win 10 Pro x86 VM that is doing everything I need it to except autologin. I can get the VM to autologin, but NOT with ENHANCED MODE active which I MUST HAVE in order to get full resolution to match my monitors etc and even more important sound.

Access to the host systems "sound" etc can only be done, for some, I am sure, VERY STUPID SHORT SIGHTED REASON, with enhanced mode active.

When enhanced mode is active the system, from what my research tells me, has the "base" VM running, somewhere in the background, and then uses RDP to open up a session in to it through which it then gains access to the host systems sound etc. What a stupid short sighted design. AT ALL TIMES A VM SHOULD ALWAYS BE ABLE TO ACCESS THE SOUND and FULL RESOLUTION of the host system, at the least, at ALL TIMES EVEN IN BASIC MODE....DUH. Think audible "problem" alerts as an example of why this is important.

In general I could live with that IF RDP would support autologin, which for obvious security reasons it at least should not do without a lot of warnings of the dangers.

I have noticed that whenever I RDP in to one of my real PCs there is an option to save my "log in" I do not get this option when Hyper-V is doing the RDP session via Enhanced Mode. (I have read that this saved "log in" can also be manually done via a settings type of file) I have actually, for security reasons, never used that option so I don't know whether it saves both username and password or just username.

My question for today is can anyone out there figure out a way to get some method of autologin to work for a VM running under Hyper-V AND also keep access to the full resolution of the system and the host systems sound?

Is there a "settings file" I could edit that could help me?

Any "thinking outside the box" solutions for this problem would also be greatly appreciated.

A VERY BIG THANK YOU to any who try to help,

Ralph

The following is the message I get:

The WS-Management service cannot process the request. The service is configured to not accept any remote shell requests.

 I configured "Allow Remote Shell access" in the GPO but it made no difference. I have since returned it to its default value.

Not that this should matter, this server has IIS, DFS and Certificate services installed. This was also an in-place upgrade from 2016. 2016 on this server had the same issue, so I did the upgrade to see if it would fix the issue and well it does not. I have another 2019 server in the domain that had no problem installing Hyper-V, so I am very doubtful this is a GPO issue. I am guessing it is a local permissions issue. I did try changing the execution policy for power-shell and that made no difference either. Again any advice would be great - Thank you.

Hello,

I'm testing Server 2019 out as a migration path but I'm having an issue with Hyper-V on where it fails to start any VM that has a vTPM enabled. I tried to enable a vTPM on a test VM, and initially it failed with the error:

TimeCreated  : 11/26/2019 8:12:15 PM
ProviderName : Microsoft-Windows-Hyper-V-VMMS
Id           : 15500
Message      : 'vtpm_test' failed to start worker process: The object was not found. (0xC0000225). (Virtual machine ID CFBC4A73-...)

Since this was a test hypervisor,  rather than troubleshoot the error I wiped it clean, and re-installed Server 2019 DataCenter from scratch, re-installed the Hyper-V role, and tried to provision another fresh VM with vTPM. However I received an error that the key protector couldn't be unwrapped:

TimeCreated  : 11/26/2019 8:45:01 PM
ProviderName : Microsoft-Windows-HostGuardianService-Client
Id           : 2014
Message      : The Host Guardian Service Client failed to unwrap a Key Protector on behalf of a calling process. This event will normally correspond to a failure to startup a shielded virtual machine. Consult the description for further details. This could be related to an attestation issue, a Key Protection Server issue, or a network connectivity issue:

               System.IO.FileNotFoundException: Could not find file 'C:\Windows\system32\config\VSMIDK'.
               File name: 'C:\Windows\system32\config\VSMIDK'
                  at Microsoft.Windows.KdsClient.HgsClient.UnwrapProtectionDescriptor(Byte[] ingressProtectionDescriptor, Byte[]& encryptedTransferKey, Byte[]& encryptedWrappingKey, Byte[]& encryptedKeys)
                  at Microsoft.Windows.KdsClient.Interop.ManagedEntry.UnwrapKeyProtector(IntPtr keyProtectorPointer, IntPtr unwrappedKpPointer, IntPtr errorContextPointer)

Doing some digging around, this file is supposed to be auto-created but it wasn't, so I just went ahead and copied it over from one of my Server 2016 installs. After that, I still receive the same error when trying to add a TPM to a VM and start it:

TimeCreated  : 11/26/2019 9:13:15 PM
ProviderName : Microsoft-Windows-Hyper-V-VMMS
Id           : 15500
Message      : 'vtpm_test' failed to start worker process: The object was not found. (0xC0000225). (Virtual machine ID ED380AFD-...)

It's worth mentioning that VMs start without a vTPM and I've also tried installing the HostGuardian feature to see if that helps. Also, I'm not experiencing this issue on a fresh Windows Server 2016 install nor on the established Windows Server 2016 boxes I have. Any thoughts as to what may be causing this issue?