I tried download new ISO from Microsoft site even though it show the same error message unabel to find the driver to install

Hi 

I've installed the Windows Server 2012 several times.

When i active the windows and install the drivers (local drivers or online updating drivers), windows works correctly. But when i install the Hyper-V Role , Win Server 2012 doesn't boot with GUI.

I went to the Safe mode and uninstalled the drivers (like Graphic Driver) but nothing changed. I think everything works fine but Win Server 2012 Does NOT boot (System restarts without seeing the microsoft Logo when it boots. )

I need to install the Win Server 2012 in my Physical System and user Hyper-V to test some roles , ....

My Physical System:

Core i5 , 4 GB RAM, 2TB HDD, Gigabyte Motherboard, 3 NICs , 1 USB port to ADSL modem, ...

I'm a little confused guys.

Hello, I've issued SAN certificates for both my primary server (domain1) and replica server (domain2) from my enterprise CA and replication with these certificates works just fine if I use the internal name of my replica server when enabling replication on a specific VM.  If I change the name of the replica server to the public address though "servername.dyndns.org"  (using dyn for some cheap external access testing) in order to try replication over the WAN then I get the following errors.  This is with the same certificates and configuration that work just fine with the local name.  I have all of the internal and external names on the certificate as well as Subject Alternative Names, I've imported the CA certificate into the trusted root store of the Primary server, have the hyper v replica open on the firewall, and am forwarding 443 through my external firewall as well.  Thanks for any help!

Enabling replication failed.

Hyper-V failed to enable replication.

Hyper-V received a digital certificate that is not valid from Replica server
servername.dyndns.org.

Hyper-V failed to enable replication for virtual machine New Virtual Machine’:
The specified certificate is self signed.

(0x80092007).
(Virtual Machine ID OBF44cBD-O8AF-4430-BSBF-FBFc3EO13F8D)

Hyper-V received a digital certificate that is not valid from the Replica server
‘servername.dyndns.org’. Error: The specified certificate is self signed. (0x80092007).

Hi

We're dealing with a strange problem regarding disk performance issue inside VM running on a 3 node Hyper-V 2.0 cluster connected to HP EVA 44xx storage system.

We analized performance on EVA, Hosts, VM and came to the following clues:

- running about 10 VM there is no perofrmance issue. All VM have dynamic disks, capacity between 50-150GB (disk R/W is between 50-70MB/s)

- running 3 virtual machines with dynamic disks capacity 400GB  or more there is a huge impact on disk performance (disk TR/W is between 10-15MB/s)

We did shut down on all VM on all nodes, and start-up only one VM that has 400GB vhd drive (inside VM there are two partitions on Windows 2008 system 50GB and data 350GB). We also stop all services running on this VM so that VM is actually idle, but disk performance is still 10-15MB/s.

This happens on all 3 VM with virtual disk 'large' vhd (about 400 or 500 MB).

On the other hand, running identical VM with 150GB vhd performance is 5x better.

HP found no problems with storage, so we're quite sure that there must be some kind of a bug or issue with VM or Hyper-V 2.0 (we're running free version of Hyper-V Server 2008 R2).

Any ideas or clues how can we proceed next? Has anyone had the same problem with Hyper-V so that can give us additional data to proceed ?

Regards,

Miha

Historicaly we have always had a stand-alone fileserver with a local tape backup drive along with making robocopy backup of the files to another remote server overnight. This server also hosts the printers for the LAN.

One issue I have found is that the Print server sometimes crashes and requires me to reboot the mail fileserver during the day, hardly ideal! (Simply restarting the print service doesn't work)

I have just purchased a new server (We can't afford SAN) for a new fileserver which will be built with Windows 2012.

Can anyone advise me on the best way to build this for the future?

I would ideally like to run a seperate VM for the print server, so this can easily be restarted (Does this make sense?)

What should I do for the fileserver to ensure some redundancy? run a 2nd VM for a File server? clustering?

I have a seperate backup server off-site, which I currently run simple robocopy jobs to overnight. If I upgrade this to Server 2012 can this be in a cluster or a on-line standyby?

I guess this all depends on what I am trying to achieve:-

a. Should I simply aim to have a 2nd server which has a current copy of the files ready to go (This is what I have from the evening before), but it wouldn't have all the shares ready to go... so would require some work to bring on-line if the primary FS went offline.

b. A 2nd server which is using DFS to have a copy of the files which would be quicker to bring on-line

c. Clustering to ensure immediate change-over, but would this work with 2-3TB of storage needed for the various network shares and files

d. Some other solution I haven't thought about?

I'm sure there must be a neat and sensible solution to my needs and it must be something other SME's also need.

Any advice or suggestions/thoughts gratefuly received.

Hi All,

Im reading this technet article when having 2 NIC in a team, that are shared for the Host (Host, CSV, LiveMigration) and for the VMs http://technet.microsoft.com/en-us/library/891c6ec4-deb9-447b-afc1-3b01151d9e4b(WS.11)#bkmk_2

I configure QoS:

# Create a NIC team that consists of two 10GbE NICs
New-NetLbfoTeam “2x10GbE Team” –TeamMembers “10GbE NIC1”,”10GbE NIC2” –TeamNicName “2x10GbE”
# Create a Hyper-V Virtual Switch that binds to the NIC team
New-VMSwitch “20GbE switch” –NetAdapterName “2x10GbE” –MinimumBandwidthMode Weight –AllowManagementOS $false
# Create a virtual NIC in the management operating system for SMB (storage)
Add-VMNetworkAdapter –ManagementOS –Name “SMB” –SwitchName “20GbE switch”
# Create a virtual NIC in the management operating system for Live Migration
Add-VMNetworkAdapter –ManagementOS –Name “LM” –SwitchName “20GbE switch”
# Create a virtual NIC in the management operating system for Cluster
Add-VMNetworkAdapter –ManagementOS –Name “Cluster” –SwitchName “20GbE switch”
# Create a virtual NIC in the management operating system for Management
Add-VMNetworkAdapter –ManagementOS –Name “Management” –SwitchName “20GbE switch”

# Assign a weight to SMB
Set-VMNetworkAdapter –ManagementOS –Name “SMB” –MinimumBandwidthWeight 40
# Assign a weight to Live Migration
Set-VMNetworkAdapter –ManagementOS –Name “LM” –MinimumBandwidthWeight 20
# Assign a weight to Cluster
Set-VMNetworkAdapter –ManagementOS –Name “Cluster” –MinimumBandwidthWeight 5
# Assign a weight to Management
Set-VMNetworkAdapter –ManagementOS –Name “Management” –MinimumBandwidthWeight 5

# Assign a weight to each VM
Set-VMNetworkAdapter –VMName * -MinimumBandwidthWeight 1

My question is regarding the last line where I set the MinimumBandwidthWeight 1 for each of the VMs. This doesnt seem right to do, if I have 200VMs on a host wont that utilize 200% of the vSwitch? And also this cant be applied if I dont have any VM yet on the server.

Can this be done on the vSwitch it self with the command: set-VMSwitch -DefaultFlowMinimumBandwidthWeight 50.

Does this mean that the entire traffic from all of the VMs that dont have MinimumBandwidthWeight defined will be in 50% (i have 100 VMs, so each VM will use 50%/100=0.5%) or each VM will get 50% (100VM, each uses 50% = 5000%)?

Thanks

Zarko

We have a new Win2008 R2 Failover cluster with 3 nodes and a set of 30+ VM's exported from a series of stand alone Hyper-V hosts. Each VM has between 6 and 30 snapshots. All hosts and VM's have the same virtual network name. When imported on a cluster node all network settings are ok and all snapshots work.

If we move or migrate a VM to another node than where it was imported, all snapshots on the VM show a configuration error on their network adapter. The guest OS has no network anymore and the snapshot config does not allow for it to be changed.

What is the point of migrating anything if it break everything that isn't running on its original host? How can I fix this?

We've been using Server2008R2 with Hyper-V for a couple of years now and chose it at the time because of its ability to make use of Intel SpeedStep and AMD PowerNow!

Now with Server 2012 and Hyper-V V3, all power management abilities seem to be gone. The CPUs are always at full speed and our servers need twice the energy as before while idling. (Yes, the CPU P-states are enabled in the BIOS)

Is this by design? Is there a workaround to enable cpu power management again? Despite the great new features of Hyper-V 3, this would be a show-stopper for us since we are very concerned about energy consumption.

rosch

I have to standalone system 

Host 1 and host 2

and I got below error

Hyper-V failed to enable replication for virtual machine 'testrep': A connection with the server could not be established (0x00002EFD). (Virtual Machine ID 4458ED53-B365-4BDB-AB27-B5D7CDC2FECB)

Hyper-V cannot connect to the specified Replica server 'Replicatest.test.com'. Error: A connection with the server could not be established (0x00002EFD). Verify that the specified server is enabled as a Replica server, allows inbound connection on port '80', and supports the same authentication scheme.

I checked every possibility according to error  

I am not able to do replication

Please Help Me

Hi All,

  We have Hyper-v cluster with 6 nodes. Currently I’m facing some issues with hyper-V cluster CSV backup with EMC networker. Current configuration as below.

All the servers are hosted in different data center . we are using vendor backup solution and it’s in different Network segment . we have DEDICATED NETWORK CARD FOR EACH HOST and added the static route for all the hyper-V servers. And hyper-v host can ping to Media server. But when the media server try to backup it’s not starting.

OS =  Windows 2008 R2 Datacenter SP 1

Network configuration :

Separate network card for all the Interfaces.

Management Network : 10.1.100.XXX range all the six servers

Cluster IP  : 10.1.100.XXX

Live migration network : 10.1.3.XXX range all the six servers

HB  network : 10.1.3.XXX range all the six servers

DMZ Network :

VM Network  :

Backup network : 10.1.113.XXX range all the six servers

EMC networker application cannot communicate to hyper-cluster cluster. any idea why? Is it possible to add new backup vlan IP address to Cluster Name?

Thanks,

Aucsna

We've got a couple of Windows 2012 / Hyper-V 3 clusters, and I have a problem with network adapters on them.

If the VM is powered off, and I move it to another host (quick migration via Failover Cluster Manager), the network adapter loses it's connection to the virtuall switch - it turns to "Configuration Error".

If it's powered on, both live migration and quick migration work fine.

It's possible to demonstrate on an empty VM. I created a new VM via the Failover Cluster Manager (ie not via Hyper-V manager on one of the nodes). One NIC, attached to the single VSwitch we have. Not much memory, disk. Don't need to install an OS.

Quick migrate with the power off, and it fails to pick up the VSwitch on the next host.

Turn it on, and live migration and quick migration both work fine.

Is this supposed to happen, or a buglet in 2012?

We have a new HP server, running Server 2012 Standard w/4 nics.  2 nics are disabled, one nic is connected to our local lan, and the other active nic is connected to our ISP only.  Both nics have static IP addresses, with the ISP nic the only one having a gateway so all traffic is routed through that nic with the exception of the local lan traffic.

The reason for the above is we have made an image of a live SQL server that we would like to boot up, install all windows updates, and test upgrading to latest version of SQL, etc... 

I have created a virtual switch, called it DMZ, and associated it with the nic that is directly connected to the ISP.  I have turned off Allow management operating system to share this network adapter.  I have also assigned a static IP to this NIC in the VM.  From within this VM, I have internet access that is going out through the ISP rather than local lan as verified by tracert.  This works and looks good.  The only issue we have is, we can still ping internal ip addresses on our local lan.  We cannot resolve names, but can ping IP addresses, which is a little scary as we want to completely isolate this from our internal network so we don't foul anything up production wise.  We have tried everything from creating false routes, etc... but after failing to ping 1 time, it someone autocorrects itself and finds the route back again.  I have scoured everything and cannot find the answer.  Any help would be much appreciated.

I have a Windows Server 2003 R2 SP2 as a Guest OS, converted using Disk2VHD,  running on Windows 2008 R2 Enterprise SP1.

Hyper-V version is  6.1.7601.17514.

When Guest OS first boots I have keyboard input because I can do a F8 and boot into safe mode or safe mode with networking.

After Guest OS boots I have no keyboard or mouse control - at all.

Other Guest OSes, one is another converted 2003 Server and one is WXPP, are running fine.

I have no access to Guest OS via RDP because I cannot get past the Windows Product Activation Window to even try and get device manager to configure legacy network adapter.

I have edited the registry and enabled autologon and that is working as it should or I would not be getting the Windows Product Activation Window. I have also edited the registry and disable the Windows Shutdown Reason and Tracker to get past that Pop-UP.

I have been unable to figure out how to tell the Windows Product Acitvation Window "No" to make it go away so that device manager might load the driver for the legacy NIC.

Any suggestions would be greatly appreciated!

I have Server 2012 Hyper V loaded on a Dell R320 with 4 500gb drives in RAID-5 on a Perc 310.

I am able to connect remotely via Windows 8.  

I can create a VM without a disk attached.

I jsut can't create a VHD.  I am trying VHD, 10gb, fixed size.  Any ideas?  

So I'm creating a new host with Server 2012 core and I'm going to run server 2012 essentials on a VM.  I'm trying to decide what to do with the disks.

I have created a 2TB volume on the physical layer that is thin provisioned using storage spaces and would like to create a 1.5TB vhd for essentials main disk but if I create this as a fixed volume then will it use the full 1.5TB of the physical volume?  If so, what's the point of thin provisioning at the physical layer for virtualized workloads?

If I used thin provisoning at the virtual layer then do I pay a performance penalty twice?  Once with the VHDX file and then again at the physical layer?

Any recommendations on how to set this up so I make the most efficient use of disk space here?

I had 5 VMs running on a Server 2008 R2 Datacenter (it was the main domain controller)

I lost the server however I was able to save the VHDs and XMLs. which I copied to another DC (same domain as previous)

On the new server I tried to create so far 2 of the VMs with the same name as originally and pointed to the existing VHD file.

The first one when I start, it shows status = running but I only see a black screen, tried Action>Insert Integration Services setup disk which completed successfully (as per status bar) but nothing, still a black screen.

The second VM, I did the same but this time I get the DOS "Boot Failure, Reboot and select proper boot device". I've tried to replace the content of the XML file with the content of the old file + changing a few lines (path and reference to its name), disabled the NIC and still same issue.

I don't want to take the chance with the next VMs until I have a solution...

Any help would be appreciated.

Im setting up the following for a client

1 Dell R710

2 Dell 2950's

1 Dell MD3200 w/ 12 7200rpm Disks in RAID 10 and dual iSCSI Modules

1 iSCSI Dedicated Dell 5448 GB 48 Port Network Switch

I plan to cluster all 3 servers and install Hyper-V Clustered on Clustered Storage to a single LUN on the MD3200 (not including Witness LUN)

I would prefer the customer change the models to SAS but that might not be a possibility

I would like to know realistically how many VM's we can run on each server with 2 iSCSI connections to the MD3200 setup in Dynamic Least Queue Depth in MPIO (or suggest a better mode)

Each iSCSI connection will connect to a separate RAID Module on the MD3200 

I know you would like to know what the virtual machines will be doing but lets just assume for now they will be a mix of standard servers, Exchange, AD, DNS, DHCP, SQL etc... just picture a nice mixture of about 30 Virtual Machines with varying disk load, nothing too heavy.

I have 4 MD3200 at my full time day job and 2 are used in 2 node SQL clusters, 1 is used for 2 Linux Nodes running Oracle and the 4th is used in a Clustered VMWare configuration. The difference is all of ours are SAS and we use all 15k drives instead of 7.2k like my current client has. Our ESXi Cluster has 3 servers and 30+ Virtual Machines and barely breaks a sweat, once I add the 4th server I plan to load anoter 20+ VM's and see no issues in the foreseeable future. It also helps that all of our SQL traffic is piped out to two SQL clusters with their own MD3200's so nothing running on our cluster has heavy disk usage.

Now my current client on the other hand will be using iSCSI and Gigabit connections which in my experience can only handle a few virtual machines per GB connection, maybe 5 depending on what they will be doing. I think 10 VM's per server running over 2 iSCSI connections will be pushing it and thats an incredible waste for the R710 which I know I can run all 30 of our virtual machines on one of our R710's with 196GB of ram and the current top of line Xeons. I dont recall what is in my clients R710 but memory can be added and as long as it has dual quads it can easily run 20 virtual machines. 

Well that's all of it. Let me know what you guys think about the iSCSI connections, any real world similar setups and experiences would be very much appreciated.

Here's my setup:

I have a wireless adapter connected to my host server which gives me Internet access.

I have one desktop connected to my host server via Ethernet, primarily for Remote Desktop access (from desktop to server, not vice versa).

It is simple to share Internet between host and desktop via ICS. Remote Desktop Connection works well too.

However, I am having trouble when introducing this next variable:

I have created a Debian VM in Hyper-V on my server, which I would like to access the Internet and use as a web host. I have created the "Internal Virtual Switch" and assigned it to the VM as legacy. For whatever reason, it will not let me assign the wireless adapter as an External Virtual Switch. I can only access the Internet on my VM if I use ICS on the wireless adapter and assign it to the Virtual Switch (Internal). This is an issue because it no longer shares Internet with my Ethernet port which connects to my desktop. I want to be able to access internet on all three machines-- desktop, server, and VM.

I have tried bridging but doing so then causes inability of connecting to the server from my desktop via Remote Desktop Connection (can't find host on network).

Any ideas? I have asked the question elsewhere and someone mentioned static IP address. My wireless internet is dynamic IP.

Hi,

I use HyperV manager on my Win7 laptop to manage varios HyperV host servers.

2 of these have since been changed and no longer exist.

But my HyperV Manager console still shows them down the left hand side with red X in obviously as it cant contact them.

If i hit Remove Server - they go away for that session but just return next time i fire up HVM!

How can i permanently remove old host machines from my HVM console please?

Thanks