Poor guest performance Server 2012 w/ iSCSI

May 1, 2013, 11:46 am

≫ Next: virtual machine is lost the network connection in windows server 2008

≪ Previous: Hyper-V Memory cannot increase on Window 2008R2

I'm not sure if this would be a better suited question for the plain old Server 2012 forums, so feel free to move this topic moderators.

Quick back story on our infrastructure. We have been building our own servers for the past couple of years to serve as a private cloud foundation using SCVMM 2012 so that our engineers can build out and deploy VMs as needed without having to ask IT to do this for them. This has been working great up until recently when I decided to use Server 2012 after the recent release of SCVMM SP1 instead of 2008 R2 on our latest build.

All 4 machines are identical other than operating system, 3 of which are using 2008 R2 Enterprise and the most recent 2012 Standard. 2 of these boxes are running 4x2TB locally in a RAID10, one of them is using the Microsoft iSCSI initiator to a Microsoft iSCSI Target we setup that has a 8TB LUN and the latest Server 2012 is also using the iSCSI initiator storage to house VHDs and VM config files.

The Server 2008 R2 box currently has 24 VMs running with the iSCSI disk reporting a highest active time of 30%. The VMs work seamlessly and I've had no complaints.
The Server 2012 box currently has 12 VMs running with the iSCSI disk jumping between a highest active time of 10% and 100% and of course when it spikes to 100% the VMs become incredibly slow and sometimes completely non-responsive. Of course I'm getting complaints left and right.

The NIC configuration on the hosts are identical since the motherboards are the same the NICs are the same. I'm trying to figure out why the IO is so terrible in 2012 but works fine in 2008 R2.

↧

virtual machine is lost the network connection in windows server 2008

May 9, 2013, 2:22 am

≫ Next: c# wmi: retrieving network adapter information when virtual machine is powered off

≪ Previous: Poor guest performance Server 2012 w/ iSCSI

virtual machine is lost the network connection in windows server 2008 not in win server 2008 R2 we are using only win server 2008

↧

c# wmi: retrieving network adapter information when virtual machine is powered off

February 16, 2012, 12:16 am

≫ Next: Hard path limit of 8 in VM with Hyper-v virtual fibre channel?

≪ Previous: virtual machine is lost the network connection in windows server 2008

Hi,

Following is the code I am using to retrieve information for network adapters attached to my virtual machine named 'XPSP3'

            ManagementScope manScope = new ManagementScope(@"\\.\root\virtualization");

            ObjectQuery queryObj = new ObjectQuery("SELECT * FROM Msvm_ComputerSystem WHERE ElementName='XPSP3'");

            ManagementObjectSearcher manSearcher = new ManagementObjectSearcher(manScope, queryObj);
            ManagementObjectCollection manCollection = manSearcher.Get();

            ManagementObject vm = null;
            foreach (ManagementObject o in manCollection)
            {
                vm = o;
                break;
            }

            ManagementObjectCollection ethernetPorts = vm.GetRelated("Msvm_SyntheticEthernetPort");

When the virtual machine is powered off 'ethernetPorts' collection is empty. And when it is running, 'ethernetPorts' contains the information about the netwrok adapters attached to the virtual machine.

How can I retrieve this information when the virtual machine is powered off?

↧

Hard path limit of 8 in VM with Hyper-v virtual fibre channel?

April 13, 2013, 9:05 am

≫ Next: How to Get Virtual Network Adapter Name of Virtual Machine using C#?

≪ Previous: c# wmi: retrieving network adapter information when virtual machine is powered off

Hello,

I am seeing a strange situation where I cannot exceed 8 paths in a vfc guest and wanted to see if anyone else has seen this.

I have server 2012 parent with Hyper-v and connecting to a storage controllers which have 16 target ports. 8 nodes, 2 ports each.

I have a brocade dual fabric and connect a single initiator to 8 targets in a zone on each fabric.

After enabling the zones I see on the parent 16 virtual ports in my qlogic qconvergedconsole on the Parent server. 8 on each HBA port. I go to my guest and I do not see a lun. I reboot the guest and check for a lun nothing.

After rebooting the server 2012, I check the vm and a lun is seen with 8 only paths. I check the parent and now I only see 4 virtual target ports per HBA.

I tried the same scenario with 4 node controllers and 4 target ports. Same behavior. 16 targets show up in HBA, after reboot of parent server only 8 ports.

Does Hyper-v have a hidden limit of only 8 paths and reduce the number of exposed virtual ports in the parent server?

Has anyone tried more than 8 paths and it work?

Thanks in advance.

↧

How to Get Virtual Network Adapter Name of Virtual Machine using C#?

May 9, 2013, 5:39 am

≫ Next: Enabling Replication Fails

≪ Previous: Hard path limit of 8 in VM with Hyper-v virtual fibre channel?

Hello ,

How to Get Virtual Network Adapter Name of Virtual Machine using C#?

Thanks

Alok kumar sharma

↧

Enabling Replication Fails

May 9, 2013, 11:11 am

≫ Next: Hyper-V + Failover Clustering + virtual network adapter failure on one node = ?

≪ Previous: How to Get Virtual Network Adapter Name of Virtual Machine using C#?

We are a vmware shop evaluating Hyper-V 2012 for the first time. I've done quite a few searches on this error, but have not been able to find a solution for this issue we are experiencing. Your help would be greatly appreciated.

We are running Hyper-V 2012 SP1 on two physical hosts with a separate machine running SCVMM and another running AD and DNS

When I use Hyper-V manager to enable replication for one of the vm sessions, I am able to get to the last part of the process, but when I click finish I am getting the following error:

Hyper-V failed to enable replication
Failed to perform the operation, the virtual machine is not in a valid state to perform the operation.
Hyper-V failed to enable replication
 for virtual machine 'vmtest1'; Operation Aborted (0x80004004)

The firewalls are disabled via GPOs,

Same error with and without snapshots on the VM

There is no media attached to the vm (CD/DVD set to none)

It does not matter if the vm is running or shut down

SPN's all check out

Using Kerberos on Port 80

Port 80 responds via telnet

The folder structure get's created on the destination (Hyper-V replica folder)

Using local disk

No NIC teaming setup

If you need any additional information just let me know. The replication piece is a critical feature for us to move forward with implementing Hyper-V

↧

Hyper-V + Failover Clustering + virtual network adapter failure on one node = ?

September 5, 2008, 8:43 am

≫ Next: Looking for a Hypver-V differential backup software in VHD level, not file level.

≪ Previous: Enabling Replication Fails

Hello all,

I have a question regarding a newly constructed server 2008 (x64) cluster running hyper-v w/ iscsi. I have two poweredge 1950 systems clustered together that have (currently) only one of the NIC ports on each server assigned to a virtual network adapter. Planned failover via simulated failure, cluster failure, and resource moving, all works as designed. However...

Is there a way for the cluster to detect a failed link for a virtual network adapter as an actual cluster failure? For example, if the following is true:

Node A -
physical adapter 1 (bound to virtual network adapter 1) - connected to switch A

Note B -
physical adapter 1 (bound to virtual network adapter 1) - connected to switch B

I simulated failure by removing the link between node A and switch A. The failure is definitely recognized under the failover cluster management MMC (the virtual network for the two adapters <vnet1> shows adapter 1 on node A as failed). However, no host migration occurs.

Basically, I want the cluster to recognize the failure of the virtual network condition on one node the same way as it views a failure of the cluster service on that node and to migrate all the virtual machines to node B.

Can this be done?

↧

Looking for a Hypver-V differential backup software in VHD level, not file level.

May 9, 2013, 9:51 am

≫ Next: Disable write cache on VHDX in Hyper-V DC guests.

≪ Previous: Hyper-V + Failover Clustering + virtual network adapter failure on one node = ?

Hi,

we are currently using Symantec Backup Exec to do file level backup on Hyper-v guests. but we would like to do an additional backup in VHD level. so we can easily revert back to previous periods. it's kind of like the snapshot, but I would like it to be differential not incremental. the idea is performing a full VHD backup during weekend and differential backup during weekdays. Then another full VHD backup on another weekend, so we have enough time to backup all the guests overnight and don't waste too much storage space since a full VHD file is huge and takes couple hours to backup.

Anyone know if there is any software like that ?

↧

Disable write cache on VHDX in Hyper-V DC guests.

November 9, 2012, 11:13 am

≫ Next: Quistion from beginner for VMs.

≪ Previous: Looking for a Hypver-V differential backup software in VHD level, not file level.

I have a couple of DC's running Server 2012 as Hyper-V guest machines. I have write caching disabled on the hosts RAID controller (no cache battery), but for some reason, I can't disable write caching on the VHDX virtual drives that the DC guest servers reside on. This has caused me issues twice during extended overnight power outages that have caused unclean shutdowns of the physical server, corrupting AD databases on the guest servers. I just spent four hours force-removing AD from one of the servers, rejoining, and fixing all the other associated issues due to corrupted AD databases.

How the heck do you disable write caching on the boot drive on a Hyper-V guest when the guest OS says the device does not allow the setting to be changed? Oh, the VHDX is mounted as an ATA device. Do I need to create a tiny VHDX to mount as an ATA device as a boot volume, and then install the OS on a SCSI mounted VHDX, or how else can I fix this problem?

Any suggestions would be greatly appreciated!

↧

Quistion from beginner for VMs.

May 9, 2013, 3:59 pm

≫ Next: Start / Stop of a vm via scheduled task fails when no user is logged on

≪ Previous: Disable write cache on VHDX in Hyper-V DC guests.

Hi,

I played with a Hyper-V before on my Windows 8 machine only and soon I am going to buy a Windows 2012 Server Standard edition for a small company and latter on install Hyper-V Servers on it to replace two very old windows 2k3 Servers.

I know there are two Hyper-V licenses come with a Standard Edition. What I do not clearly understand how many Windows Servers will I have with it - two or three?

In other words can I Install DC, DNS, and DHCP on a Host and then setup two Virtual Servers and install File Server on one of them and Exchange on other. How many network cards should I have?

Any explanation and suggestion would be appreciated. We do have a very tight budget.

Thank you in advance,

Alex

↧

Start / Stop of a vm via scheduled task fails when no user is logged on

May 9, 2013, 7:25 am

≫ Next: Can't RDP to Virtual Machine from Host

≪ Previous: Quistion from beginner for VMs.

Hello,

I want to start / stop my two hyper-v virtual machines in a scheduled task. I use Windows Server 2012.

For this purpose I created two cmd-scripts which calls a powershell-script.

vor64.cmd:
chcp 1252
C:\Windows\System32\WindowsPowerShell\v1.0\Powershell.exe -noninteractive -nologo -command "& 'c:\batches\vor64.ps1'"

vor64.ps1
stop-vm -name "vm1" -force
stop-vm -name "vm2" -force

The nach64.cmd/nach64.ps1 looks similar only stop-vm is replaced with start-vm.

The scheduled task are konfigure to run under the administrator-account.

The scripts work fine if I start them via explorer: vm stops and start again.

They will although run fine as a scheduled task if the administrator is logged on while the task is executed. But they will not work if no user is logged on the server console.

Why?

Stefan Westner

↧

Can't RDP to Virtual Machine from Host

May 9, 2013, 6:31 pm

≫ Next: no internal connection between 2 VMs

≪ Previous: Start / Stop of a vm via scheduled task fails when no user is logged on

I am new to the Hyper-V world and have come across a roadblock. I have Server 2008 SP1 with current updates installed and have added the following roles:

Hyper-V, Network Policy and Access Services, Terminal Services and Web Server (IIS)

My host machine, Server 2008, has a static IP assigined. My virtual machine is getting it's IP from the DSL modem that I am using and is in a different network (192.168.x.x)

My goal is to have the ability to remote desktop to the virtual machine without interfacing with the server. I would like to be able to do this with username and password authentication.

I have tried every combination of network adaptor on the virtural machine and have not been able to access it via remote desktop from my host machine.

Is what I'm trying to do even possible with out assigning a static IP to the virtual machine?

Thank you in advance for any help.

↧

no internal connection between 2 VMs

May 9, 2013, 8:09 pm

≫ Next: Hyper V 2012 Backup - best practice

≪ Previous: Can't RDP to Virtual Machine from Host

I am using windows 8 hyper-v.

I added an internal virtual switch for hyper-v, assigned IP 192.168.1.100 to it.

Then I created 2 VM, assigned 192.168.1.101 with gateway IP above to one VM, and 192.168.1.102 to another.

VM couldping 192.168.1.100 but for VM I could not ping each other..., no internal connection between 2 VMs

↧

Hyper V 2012 Backup - best practice

January 2, 2013, 9:00 am

≫ Next: Comment donner au service Hyper-V l'accès à une image ISO sur un NAS ?

≪ Previous: no internal connection between 2 VMs

I'm setting up a new Hyper-V server 2012 and I'm trying to understand how to best configure it for backups and restores usingWindows Server Backup. I´ve 5 VMs running, everyone has their one iScsi drive.

I would like to backup the VMs on a external usb disk. I´ve the choice to backup the Hyper-V using child partition snapshot or to backup the vm folder with the config and the vhd files (is is right, that Windows override the old backup ?)

What´s the difference between the two methods ? Is incremental back possible with vm´s ?

thanks

↧

Comment donner au service Hyper-V l'accès à une image ISO sur un NAS ?

May 10, 2013, 2:29 am

≫ Next: VMC not refreshing

≪ Previous: Hyper V 2012 Backup - best practice

Bonjour,

Avec Hyper-V sur Windows 8, impossible de créer une nouvelle machine virtuel si l'ISO indiqué pour installer le système d'exploitation se trouve sur un serveur externe.

J'ai pourtant lancé Hyper-V en tant qu'administrateur et le service "Gestion d’ordinateurs virtuels Hyper-V" est exécuté avec le comte système local, qui a un accès total à mon NAS.

Voici l'erreur affichée :

Echec de l'ajout du périphérique 'Virtual CD/DVD Disk'.

Le compte d'ordinateur "Service de gestion d'ordinateurs virtuels Hyper-V" et/ou l'utilisateur initiant l'opération de gestion de l'ordinateur virtuel ne disposent pas des autorisations d'accès requises au partage de fichiers \\NAS\...\MonFichier.iso".

Merci pour votre aide.

↧

VMC not refreshing

May 2, 2013, 8:51 am

≫ Next: Virtual Network Name on Multiple Hosts

≪ Previous: Comment donner au service Hyper-V l'accès à une image ISO sur un NAS ?

Hi. I use Windows Server 2012 and many VM's on this server.

And now i noticed, sometimes if i logn time not working with VM - display this VM is "pused". As example:
i run Windows Update and wait when it finish, but Windows Update paused. I still wait when finish install 2 updates - but long time never happens. And process-line have one position. When I get tired of waiting i go to VM and display switch to actually moment )

Is it possible to fix the display refresh? Because I can wait for hours (

Thank you!

↧

Virtual Network Name on Multiple Hosts

February 9, 2012, 1:37 pm

≫ Next: Unable to Add New Virtual Switch using SCVMM SP1 to Windows 2012 Datacenter Server Hyper-V Hosts

≪ Previous: VMC not refreshing

Hello All,

Will naming the virtual network name the same on two different hyper-v hosts cause any issues if failover is not configured?

Is it best practice to name the virtual networks different or the same?

I see where if failover the virtual network name needs to be the same.

Thanks in advance.

↧

Unable to Add New Virtual Switch using SCVMM SP1 to Windows 2012 Datacenter Server Hyper-V Hosts

May 10, 2013, 4:02 am

≫ Next: Windows Server 2012 - Can occasionally not access second virtual hard drive inside a VM

≪ Previous: Virtual Network Name on Multiple Hosts

We've even followed the examples here by Keith

http://blogs.technet.com/b/keithmayer/archive/2013/04/11/configure-network-fabric-and-network-virtualization-using-system-center-2012-sp1-vmm-build-your-private-cloud.aspx

and it does not work.

Error (505)
Virtual Machine Manager was unable to create a new virtual switch New Virtual Switch0.
Unknown error (0x80041001)

Recommended Action
Check the virtual switch name, and then try the operation again.

Any ideas?

Andrew Hancock

↧

Windows Server 2012 - Can occasionally not access second virtual hard drive inside a VM

October 20, 2012, 1:57 pm

≫ Next: Unable to create filesystem (mkfs.ext4) on large > 2TB GPT virtual disk using Linux VM.

≪ Previous: Unable to Add New Virtual Switch using SCVMM SP1 to Windows 2012 Datacenter Server Hyper-V Hosts

I run Windows Server 2012 RTM Hyper-V and I can occasionally not access the second virtual hard drive (dynamically expanding VHDX) attached to the VM through the virtual SCSI controller. I can however access the first hard drive that is connected with the virtual IDE controller.

I get the following warning in the event log under “Administrative Events” every 30 seconds when this happens:

Log Name: System
Source: Storvsc
ID: Storvsc
Message: “Reset to device, \Device\RaidPort0, was issued.”

I get this error once or twice a week and it has caused serious problems since one of the virtual servers that have this problem is a fileserver and the second hard drive contains all the data.

The only quick solution to the problem that I have found is to force the virtual machine to stop using the “Turn Off” feature since a normal shut down does not work (stops at shutting down the event log or similar) and then start the virtual machine again.

You can also wait for about 30 minutes or longer until the disk for some reason becomes accessible again by itself.

My research into this problem shows that:

Only 2 of the 10 VMs running Windows Server 2012 RTM that I have, have this problem.
Both these VMs have a second virtual hard drive (dynamically expanding VHDX) that cannot be accessed for 30 minutes or longer.
Check Disk of the virtual hard drive shows no errors.
The second hard drive is attached to the virtual SCSI controller.
I can find no problems at all with the physical storage on the (not related) 4 hosts that I have. The problem exists only in the VMs.

I have now attached the second virtual hard drive to the virtual IDE controller to see if this permanently fixes this problem (i.e. does not happen for at least a week).

Is there something wrong with the virtual SCSI controller or the virtual SCSI device driver that comes with Windows Server 2012 RTM? Does anyone else have this problem?

↧

Unable to create filesystem (mkfs.ext4) on large > 2TB GPT virtual disk using Linux VM.

April 22, 2013, 12:28 pm

≫ Next: Connection Manager opening minimized

≪ Previous: Windows Server 2012 - Can occasionally not access second virtual hard drive inside a VM

I am unable to create a file system on a large (> 2TB disk) virtual disk for a Linux VM. I can create the disk, attach it to the VM, partition it with "parted", but I cannot run mkfs.ext4. Details below.

Hyper-V 2012 Core (w/ all Windows/Microsoft updates as of 4/19).

CentOS 6.4 VM w/ 4 virtual processors, 4GB RAM, and 3 dynamic drives:

/dev/sda 100GB IDE dynamic vhdx
/dev/sdb 75GB IDE dynamic vhdx
/dev/sdc 10TB SCSI dynamic vhdx

Using parted, created 500GB partition on the 10TB drive (/dev/sdc1).

(parted) select /dev/sdc
Using /dev/sdc
(parted) print
Model: Msft Virtual Disk (scsi)
Disk /dev/sdc: 11.0TB
Sector size (logical/physical): 512B/4096B
Partition Table: gpt

Number Start End Size File system Name Flags
1 1049kB 500GB 500GB production_archive

then run: mkfs.ext4 /dev/sdc1

repeating error on console from mkfs.ext4:

INFO: task mkfs.ext4:2581 blocked for more than 120 seconds
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.

Runaway error in var/log/messages until my /var system filled up - 25G worth of it:

-rw-------. 1 root root 25085329408 Apr 19 23:15 messages

Apr 19 17:39:28 nfs2 kernel: sd 4:0:0:0: [sdc] Sense Key : No Sense [current]
Apr 19 17:39:28 nfs2 kernel: sd 4:0:0:0: [sdc] Add. Sense: No additional sense information
Apr 19 17:39:28 nfs2 kernel: hv_storvsc vmbus_0_13: cmd 0x93 scsi status 0x2 srb status 0x6

Same problem happens when running "mkfs.ext4 -E lazy_itable_init=1 /dev/sdc1"

↧

Latest Images