Hi,

I have 2 nodes cluster for my hyper-v environement. each of this nodes run multiple VMs.

I have other servers that are not powerful enough to run as many VMs at the same time. These servers could run probably only 1 VM each.

Is it possible to add these servers to my cluster, and assign only 1 VM to each one of these (low end) hosts. And if there is an issue and a  failover happens , I would like these VMs to go back to the big hosts, and then move them back to the low end host when the issue is resolved.

Basically, I want to take advantage of these small servers and run 1 VM on each, but if there is a failover, I don't want the cluster to try to push more VMs to these small servers because it won't be able to handle it. (Having a way to tell the cluster what host each VM could run on)

Thank you.

-Mehdi

I have been trying to use Virtual SANs/FC Adapters in my 2012 Hyper-v failover clusters. I have run into a couple problems which have led me to remove failover clustering and go with stand-alone hosts, which seems more stable. If anyone has experienced similar issues and has better fixes, that would be fantastic.   

SAN Details:
- HP EVA4000
- OpenIndiana ZFS storage server
- HP Brocade 4 and 8 Gb FC switches

Host details:
- HP BL465c G7 blade servers
- HP LPe1105 HBAs
- Latest HP MPIO DSM for EVA, Microsoft DSM for OI LUNs
- Latest HP SPP (firmware and drivers)

Issues:

1. If a node is reset or crashes, a VM using virtual FC adapters will not start on any nodes in the cluster--I receive "an unknown NPIV error occurred”. In order to start the VM, I have to swap the WWPNs on the A and B ports on each virtual FC adapter. The VM will not Live Migrate (since I can't use the second WWPN) until all nodes in the cluster have been rebooted.
2. After removing a node from the failover cluster, the previously working vSANs stop working. If I remove the vSANs, I cannot add them back because Hyper-v decides that the HBA driver no longer supports NPIV. The only fix I have found is a rebuild...which is hardly a fix. Haven't been able to get deep enough into HP's support to find someone who knows what "NPIV" means
3. This seems like it may be on my storage side, but a couple time my clusters started failing roles back and forth until they locked up and had to be totally rebooted to restore them. 

Matthew

We have two identical Dell PowerEdge T410's running Windows Server 2012 Standard running as Hyper-V Hosts (and nothing else).

The Hyper-V guests on one of the machines take over 40 minutes to boot.  Once booted the VM's seem to perform well.  The second T410 is running without any problems.

The Dell has 40Gb RAM and has 12 virtuals, the 5 XP virtuals are running with 1 GB RAM each and the 7 Windows 7 VM's have 2 Gb.

When everything is running, the Dell's task manager says it's using 22.9 of 40Gb RAM and the CPU is ticking over at 1 or 2%.

I've just wiped and re-installed the host with a clean Windows Server 2012 with no improvement.

Dell's technical support has had a quick look and can't see anything wrong with the hardware side of things.  The BIOS settings in the problematic machine has exactly the same settings as the BIOS in the machine that is working OK.

The clients on the 'slow' host are all part of a domain with the domain controller on the 'good' host.  Tracert and ipconfig all report as expected.

A 'Google' has turned up a couple of articles which seem to only refer to Windows Server 2008 R2, however, the problem only started to appear when I installed Server 2012.

So, in a nutshell, two identical servers, the guests on one are working fine, the guests on the other take over 40 minutes to boot.

Could anyone give me any pointers?

Richard

For me Storage Migration and Shared Nothing Live Migration.. both seems to be identical. Below are the points which lead me to that conclusion.

1) In the both we don't need any downtime

2) In the both storage will be migrated in the Non-Clustered or in Clustered Environment

3) Almost all the options will be the same except "Authentication Protocols, Incoming Live Migration and Number of Live Migrations" in the Shared Nothing Live Migration.

I would like to where is the difference.

Apart from the above again what is the difference between "Shared Nothing Live Migration" and "Hyper-V replica" ?

This is odd. I have an Ubuntu Server guest VM running in Hyper-V 2008. Samba is set up and configured with a NetBIOS name, and I enabled NetBIOS over TCP/IP on the Hyper-V network interface. With these settings, my guest VM can ping both the Hyper-V host and any other machine on my network by hostname. However, only the Hyper-V host can ping my guest by hostname. Other machines on my network can ping the guest by IP, but not by hostname. While I would think that this is a Samba/Linux issue, I have the same configuration running under a VMware environment, and everything works fine, so I'm thinking that this is an issue with my Hyper-V host config.

Any ideas?

Symptom

You use Hyper-V on a computer that is running Windows Server 2008 or Windows Server 2008 R2.

• In Hyper-V Manager, you try to try to export a virtual machine to a network share.
• You have the full control permissions on the network share.

In this scenario the export may fail with an error that resembles the following:

An error occurred while attempting to export the virtual machine.

Failed to copy file during export.

Failed to copy file from ‘<source path of VHD file>’ to ‘<network share>’: General access denied error (0x80070005)

Cause

When you export a virtual machine in Hyper-V manager, it is the System account of the Hyper-V host that executes the export. This problem occurs because the Hyper-V host does not have permission on the network share

Resolution

Ensure the permissions allow the Computer account of the Hyper-V host performing the Export to update the shared folder.

Note: While following steps are specific to 2008, the idea is the same for updating the share and NTFS permissions if the share is hosted on other versions of Windows. You can adjust the steps accordingly.

Updating the NTFS level permissions:

1.  On the destination server, Right-click on the shared folder and select Properties. 
2.  Select the Security tab. 
3.  Click Edit button and click the Add button in the permissions dialog box. 
4.  Click Oject Types and select Computers if not already done and click OK.
5.  In the dialog "Enter the object names to select" provide the name of Hyper-V host machine and click Check Names.
6.  With the Hyper-V host machine name select, click Allow by "Full control" and click OK.

Updating the share level permissions:

1.  On the destination server, Right-click on the shared folder and select Properties. 
2.  Select the Sharing tab. 
3.  Click Advanced Sharing...
4.  On the Advanced Sharing dialog, click Permissions  and then Add.

5. Click Oject Types and select Computers if not already done and click OK.
6.  In the dialog "Enter the object names to select" provide the name of Hyper-V host machine and click Check Names.
7.  With the Hyper-V host machine name select, click Allow by "Full control" and click OK.

Applies to

Running an HP DL380 G5 single quad core with 8GB of ram.

OS is on RAID-1 146GB 10k rpm SAS

VM's boot from EqualLogic SAN on 24 900GB 10k rpm SAS via iSCSI at 1Gb/sec (as a test lab)

Host OS is Server 2012 with hyper-v role (all updates)

Client VM's are all server 2012

I have been running bandwidth tests between VM's on the same node and have noticed that no matter what I do I cannot get the vSwitch to push more than about 2.0Gb/sec between VM's. I am using iperf as a test to generate bandwidth.

I have tried between 2 VM's, I have tried 3 VM's pushing to 1 VM, I have tried 2 to 2, but no matter how I structure the tests the overall aggregate bandwidth between all the VM's never exceeds ~2Gb/sec over the vSwitch.

Suppoesdly the vSwitch is a virtual 10Gb/sec adapter so I'm wondering why I can only push 20% of that traffic. I'm using iperf because it doesn't require disk I/O to test bandwidth because I want to eliminate the client OS or the SAN as the bottleneck.

Can anyone provide help on getting more performance out of the vSwitch?

Dear all,

I have a big problem with my hyper-v server and network card. I've found a lot of topic around this question but none of them helped me.

I have a Win 2008 R2 Server with only Hyper-V installed. I have 3 NIC , one is reserved to the host management and the others two are reserved for virtual machine .

All NIC are Intel, one is integrated on the Intel motherboard and the others are Intel Gigabit ET Dual port .

Before installing the role of Hyper-V all NIC worked perfectly with speed around 1Gb (file transfer average 90 MB/Sec, of course I have a Gigabit LAN) .

After Hyper-V install, the host system  is going very slow and the network file transfer using the dedicated NIC is about ten time slower (9-10MB /sec) . Same problem with all VM off and also if I remove all VM !

I tried all of suggestion found in old post (disabling TCP checksum / offload ...) but nothing happens.

What is really strange is that by removing all network card all the Host operating system seems to improve the speed... but, of course, I can't use  my LAN :(

I need help because every night I must make a full backup of Virtual hard disk from the host to a backup server in my lan.

My configuration:

- Server Win 2008 R2 SP1 with Hyper-V and no others role

- 2 Virtual Machine (one Win 2008 R2 and one Win 7 Ultimate)

- NAS Linux (access via SMB with no problem)

- Various Windows 7 Pro client

Thanks in advance to everybody,

Luca M.

I have a question concerning creating a Linux virtual machine.

I'm using a dynamic hard disk type.

Also, I have installed CentOS 6.3. The current VHD size is 9GB.

I want to know how I can decrease the VHD size.

Can someone please tell me.

Thank you in advance.

around kyushu

Dear All,

please help to configure best NIC Teaming for my Hyper V Server 2012....

4 LAN ports 1 Gbps each and i am going to NIC Teaming in below pattern :-

LAN 1 + LAN 2 = NIC TEAM A > Virtual Network Switch A > For Exchange Server 2013 VM & for DFS File Server VM

LAN 3 + LAN 4 = NIC TEAM B > Virtual Network Switch B > For SQL Server 2012 VM & for ERP Application Server VM

I have two Physical switch and i will Select teaming mode is "Switch Independent"

Physical Switch 1 = LAN 1 Cable + LAN 3 Cable

Physical Switch 2 = LAN 2 Cable + LAN 4 Cable

Question 1 :- Is above configuration is good or i need to change it? My objective is, if any of the Physical Switch goes down my users are able to communicate with servers. Also wants to utilize all the LAN ports for my VMs to balance the network load. 

Question 2 :- What should be my Load Balancing Mode "Hyper V Switch Port OR Address Hash"?

OR

Any other best option kindly suggest.

Thanks for Help!

Regards..........

We are using two Windows 2012 Hyper-v servers gui on iscsi. We have two servers connecting to the iscsi device it's a iomega ix4-200d nas. We have noticed that if on one server you make directory name change or a file name change it doesn't reflect that change on the other server... How can we fix this?

Thanks,

Mike

Hi I have W8 Pro and a W7Pro with Hyper-V.

I have a two snapshots with the virtual machine. I don't have any backup of it. 

Everything was working fine until I attached VHD using Disk Management to explore the hard drive of the W7Pro. I didn't make any changes and then just dettached it.

Later when I tried to turn on the virtual machineW7Pro I got the following error:

"Failed to Power on with Error 'The chain of virtual hard disks is inaccessible. There was an error opening a virtual hard disk further up the chain.'."

Image here: https://skydrive.live.com/redir?resid=3BC47CE0EE20515D!824&authkey=!AAw8dawZrIsxS2k

Does anyone know what happened? and how can I fix it?

Thanks

hi all,

This is my scenario:

172.16.0.0/24----windowserver2008std  (RRAS +hyperV role)-----192.168.0.0/24

                                               |

                                               |

                    hyper-V server (2008entR2-sp1) 192.168.0.1

Client from network 192.168.0.0 can access sql-server very good, but not for client from 172.16.0.0

I can ping 192.168.0.1( sql server-hyperv) , also find open port 1433 on 192.168.0.1 with port scanner from client on 172.

I tried to install ftp server on this hyperv server then get the same problem, works well with client from 192.168.0.0 , failed for 172

any idea for me? tks alot

trying to backup VM run SQl server 2008 in cluster enviroment on Windows 2012 HyperV Host. Using windnows backup utility backing upda HyperV companent and VM failed wit the follwoing error: event id 517:

The backup operation
that started at '‎2012‎-‎10‎-‎25T22:29:46.631794000Z'
has failed with following error code '0x80780049' (None of the items included
in backup were backed up.). Please review the event details for a solution, and
then rerun the backup operation once the issue is resolved.

detail info from Windows Backup utlity:

>>>>

Writer Failures
Writer Id: {66841CD4-6DED-4F4B-8F17-FD23F8DDC3DE}
Instance Id: {4DC1EF04-0A7E-4223-91D6-2A831F52DCFE}
Writer Name: Microsoft Hyper-V VSS Writer
Writer State: 0
Failure Result: 81000112
Application Result: 0
Application Message: (null)
   Component: Host Component
   Logical Path:
   Component Result: 8100010F
   Component Message: Component reports path on volume which has been excluded. (0x8100010F)
   Component: 692CFC20-574A-4B26-8A33-9FD4BFA767CA
   Logical Path: (null)
   Component Result: 800423F3
   Component Message: (null)
       File Spec: D:\VMLive\SQL Cluster Node Two\Virtual Machines\\692CFC20-574A-4B26-8A33-9FD4BFA767CA.xml Recursive: 0
       File Spec: D:\VMLive\SQL Cluster Node Two\\SQL Cluster Node Two.vhd Recursive: 0
*-----------------------------*

Application backup
Writer Id: {66841CD4-6DED-4F4B-8F17-FD23F8DDC3DE}
   Component: 692CFC20-574A-4B26-8A33-9FD4BFA767CA
   Caption     : Backup Using Child Partition Snapshot\SQL Cluster Node Two
   Logical Path:
   Error           : 80780175
   Error Message   : Component was skipped from volume shadow copy.

   Detailed Error  : 800423F3
   Detailed Error Message : The writer experienced a transient error.  If the backup process is retried,
the error may not reoccur.

Writer Id: {66841CD4-6DED-4F4B-8F17-FD23F8DDC3DE}
   Component: Host Component
   Caption     : Host Component
   Logical Path:
   Error           : 80780175
   Error Message   : Component was skipped from volume shadow copy.

   Detailed Error  : 8100010F
   Detailed Error Message : Component reports path on volume which has been excluded.

*-----------------------------*

>>>>

thanks

Hussainn

Does anyone know of a table or matrix to lookup which hyper-v features can be enabled and still do Live Migration between different Intel processors?  The servers below are available for a test lab to build a simple failover cluster, it won't be used for anything important.  When I tested something like this a couple years ago, Hyper-V management console in 2008 R2 would perform the LM just fine, but SCVMM 2008 R2 wouldn't allow it. This time we're planning to use Server 2012 and VMM 2012 SP1, we just don't want to spend time building the cluster and VMs and then find out we can't do live migrations at all.

The result below is from the cluster validation report.

Validate Matching Processor Manufacturers

The processor version is not consistent across all the nodes. This is important because Hyper-V may not be able to live migrate a virtual machine between processors of different versions.

Buenas noches a todos.

Mi pregunta es la siguiente. Tengo un servidor bastante potente el cual virtualiza tan solo 3 máquinas virtuales. Hay recursos sin problemas en el host y en las máquinas Virtuales. VEo que la comuninicacioón entre las máquinas virtuales no es la esperada, osea, si comparto un fichero entre una y otra la transferencia de ese fichero debería ser casi instantánea, ¿ No es así ?, ambas máquinas virtuales residen en el mismo servidor.

Si comunico ambas a través de la red interna de la empresa , ¿no debería influir que ambas máquinas estén dentro del mismo servidor?, ¿no debería ser la comunicación instantánea?. Hay alguna configuración especial a nivel de tarjeta de red, o procesadores (tienen 4 procesadores lógicos ambas máquinas virtuales), algo que haga que esa comunicación entre ellas sea mucho más rápida?

Muchas Gracias

David

David

Hi,

I am having some issues with a new Hyper-V Server i've just built, spec below:

Intel Core i7 2.8Ghz (Quad Core)

24Gb of DDRIII

2x 74Gb Western Digital @ 10,000rpm (SATA in RAID1)

4x 2Tb Western Digital @ 7,2000rpm (SATA in RAID5)

1x 1Gbps Quad NIC (Teamed)

Running MS Windows Server 2008 R2 (all updates\patched\service packed!)

I am finding them starting up the VM's they take a fair while to load, im running 5 VM's:

2x DC's

1x Exchange

1x Management

1x File

Any suggestions

(please note, this is a custom built server!)

I'm close, but not close enough.  Bear with me here, I'll try to be concise.

I have a Server 2012 Hyper-V host with Server 2008 R2 guest VMs. Integration services are installed and up to date, I believe. 

In the physical machine I have a SATA drive carrier with a pair of removable SATA drive caddies.  I want to be able to hot swap the 2 drives in and out periodically and have one of the guest machines see the drives and not get confused about which drive is installed at any given time.  I do not want to have to reboot the guest machine after a swap.  Can this be done?  (I don't mind if there is some process I have to run on the guest to refresh the drives as long as I can subsequently automate that process--I'll reboot if I have to.)

As close as I've come is this:  I have given the Guest OS a SCSI controller, and have assigned one of the physical SATA drives to it (call it Drive 1).  Then I boot the guest.  When the Guest boots, it sees the drive OK.  In the guest, I have assigned that Drive1 to be G:.  Then, I swap the drive and reboot the guest.  After rebooting, the guest shows the second drive.  I now assign Drive2 to be H:.   With this configuration, if I swap the drive and reboot, the guest OS shows Drive1 as G: and Drive 2 as H: depending on which drive is physically installed and it does this reliably every time.

But if I don't reboot, I get untrustworthy results.  For example, if I write something to Drive1 (G:): and then swap in Drive2 without rebooting, the guest acts as if the swap did not occur. The OS still shows the mounted drive as Drive1 and appears to let me write data to it without error messages.  But I don't trust what I am seeing because it still shows the contents of Drive1 that is no longer in the system, and also shows the new content I just added with Drive2 actually in the system.   And sure enough, if I reboot the guest, the OS now shows Drive2 (H:) and its contents without the content I added before the reboot.

Any hope for what I want to do?

Thanks,

Peter

I have tried re-running the Anywhere Access but it fails on 2 things.

1. It says 2 routers on the network may prevent port forwarding etc.

2. Also says Anywhere Access may be blocked to your server on ports 443 and 80.

My second router is a Netgear FVS318g it's kinda old.