As of sudden I have on one VM with guest Windows Server 2008 R2 SP1+ an additonal APIPA assigned for IPv4 and can not get rid of it.
Already deleted Network Adapter, remove virtual Adapter from VM Settings, removed registry Settings, but the APIPA always Comes back.
It is just on this VM which is acting as a RDP Server.
Any idea?
/Tom
Hi,
My Hyper-V installation on 2012 Data Centre used two teamed NICs as a Hyper-V virtual switch configured to connect to the external network.
After upgrading to 2012 R2 Data Centre I no longer have connectivity to the external network. VMs that use this switch no longer have network connectivity.
I've removed the switch and reconfigured but with no joy. I've also completely removed the Hyper-V role and reinstalled but again - no joy.
For the time being the NICs are configured as separate switches and all is OK.
Any ideas?
Many thanks,
Sean
After installing Hyper-V I have experienced stopped services. After I had installed guest servers and run the, this error have disappeared. Interestingly these services are stopped when guest servers go down. Here is picture from Server Manager:
Regards
Milos
Hi,
I have to emulate my physical COM1 (with installed a SIEMENS GSM MODEM), on my VM guest.
What i have to do? What is the "named pipe"? The COM must be the 1, my software only recognize modem on COM1.
Thank you all
Matthew
Hi Forum
I have this strange issue
I'M running Hyper-V on a Server 2012 (Fujitsu TX 300 S7)
with LSI SAS 6G 5/6 512MB (D2616)
I have a very large VHDX-file (around 3 TB) when I try
to make a off-line compact of the disk it Crashes with a
BAD_POOL_CALLER 0x00000c2 vhdmp.sys
Does anyone in the forum have a solution on this issue
Thanks in advance
B.Hansen
Bonjour
J'aimerais que excel et word 2007 soit les programmes par défaut
J'ai office 2010 d'installé sur mon ordi (simplement pour avoir outlook 2010 ...pour le moment)
et je ne veux pas que excel et word 2010 (je veux pouvoir les utliliser seulement sur demande)
soit les programmes
qui ouvrent mes fichiers .xls et .doc ...etc par défaut ..
Je préfère utiliser mon word et excel 2007 comme programme par défaut car
je préfère la barre d'outils et autres ...du office 2007
Comment faire SVP ??
Aussi
Mon outlook 2010 ne parvient pas à se connecter sur mon adresse courriel hotmail ...
pourquoi et que dois-je faire ??
Merci à l'avance de vos conseils ...
For non Unix/Linux folk, an equivalent would be looking at taskman and seeing 25% CPU all the time on a quad core system, yet no processes at 25%.
Dynamic Memory started working in R1 and CentOS 6.4 but there was a 'bug' where an idle system would report 1.00 for system load instead of 0.00 when it was turned on. With R2 dynamic memory is now working for Ubuntu 13.04 and Centos 6.4 but both have this same issue where an idle system reports 1.00 as baseline.
It appears it may be a 'cosmetic' issue as I don't see a process in guest or host spinning. What concerns me (besides SNMP showing a erroneous system load) is that processes that are waiting for 'idle' to do maintenance or accept more connections may behave incorrectly.
More background:
This appears to happen after memory trimming starts.
It does not appear to be a true load, no CPU core is spinning on any process I can see.
You can see the load with 'w' or 'top' or 'uptime'. Top shows 1.00 but no process with any CPU load over 0%.
I have not tried it with a single 'cpu' assigned.
When dynamic memory is unchecked and guest started the idle CPU remains 0.00
Host CPU type does not appear to matter, tested on Q9550s, Xeon 5300,5400, 5500 series.
For CentOS happened in 2012 R1 and R2
Latest kernel updates tried for both guest operating systems
-John Spade (Connect)
When I create a private network on Server 2008 r2 with the Hyper V role (hotfix for 2012 guest vms installed). I then install server 2012 as a guest and install the dhcp,wds roles. Then MDT 2012 u1. When I perform a deployment using multicast, booting into a vm pxe boot, the performance is about 600-700kbps. The network utilization is at 1-10%. A 13Gb image takes ages. If I put the legacy nic into the mdt server and of course the client as it pxe boots the performance is 10000kbps.
This is the same on a external network. At first I thought it was a physical switch issue. When I created a MDT server on a physical server (1gbps port) and turned on flow control the multicast performance is great 10000kbps.
There seems to be a bottleneck in the hyper v switch as the mdt server as its conenction is 10gbps or a lot higher than 100mbps.
Can I configure the hyper v switch for flow control (or some other way)? When using the hyper v vm nic their are a lot of NACK errors and repair packets. Their are hardly any errors when all adapters are at 100mbps.
Just wondered if anyone else had encountered this issue.
I'm currently testing with a server 2008 r2 mdt vm guest.
I'm trying to figure out how to do a VM hot copy of individual running VMs and would appreciate any insight or suggestions.
Goal
To be able to make copies of running virtual machine VHD files to another location. We want to be able to have copies of the VHD files themselves of the running systems available on another failover system. The idea is that if the single blade chassis/network/SAN go offline we have all of the VHDs in another location that we can quickly bring online without having to do a long VHD retrieval process but only what data content has changed since the VHD was last backed up.
Enviroment
Windows 2008 R2 Datacenter running on several blades in 1 blade chassis (that's all we have)
Hyper-V VMs running in a cluster between the blades
SAN attached as a quorum drive to the cluster (we only have 1 SAN)
A large Dell server with lots of RAM/CPUs/HDD space to be the failover machine (it's what we could afford)
Ideas
The problem is 90% of these VMs need to stay online so I want to be able to hot copy the running VHD files.
VMM cloning requires a VM to be offline which is not possible as does copy/paste or robocopy.
Backup software like DPM & BackupExec backup the entire VM Host server in one giant backup but I want a direct copy of the VHD itself.
Windows Backup Services makes a VHD of the device being backed up but these are non-bootable VHD files.
We've looked into Symantec System Recovery as it is what we use on our physical systems but that requires doing a drive clone image and then converting to VHD and is time consuming. It's also rather expensive even for educational licenses.
I've looked for hot copy tools for Hyper-V VHD files and I've had no luck finding good resources.
Suggestions? I'm not looking to be told to do it some other specific way, this is what I'm trying to accomplish so I'd appreciate only comments that help me attain my goal and not something like "just backup the entire host with DPM and deal with it."
Hi,
for a relatively small environment I have the following setup:
The switch (as of yet a Cisco 3750) is configured with static MAC and arp entries for the NLB multicast to work properly without flooding the network. Initially the Hyper-V hosts had only one connection to the internal network. As the MAC addresses of the Virtual IP's from both NLB clusters are registered on that port, all VM's saw the traffic to them. Therefore I added another port to the internal network, created another Virtual Switch in the internal network and connected the NLB vm's to that (and updating the static MACs of course). This way the non-NLB VM's don't 'see' the NLB traffic anymore. However, both NLB clusters are still seeing each others network traffic. For the AppV server that's not so much of an issue but the TMG's logfile is completely filling up with bogus traffic that is actually for the AppV NLB cluster. I can filter those messages out, but I'd rather not have that unneeded traffic seen at all.
I could add another physical NIC port, create another Virtual Switch and split traffic from both NLB clusters that way, but unfortunately there are no more free ports on the Cisco. Isn't there another way to split this? I cannot create more than one virtual switch on a physical port. So what I basically need is a programmable virtual switch that allows me to create something like static MAC to a virtual port - and allow me to connect a VM to that specific virtual port, in order to allow me to split the traffic.
Is something like that doable with Hyper-V 2012? Would I need some specific Virtual Switch extensions?
In Hyper-V 2012 when you create a VM the folders are created automatically into the Virtual Machine folder but are only identified by the ID of the VM.
Is there a way of automatically creating a folder with the VM name first then the VM files located in this folder when creating new VM's.
Hi
In my win 8 hyper-v test farm I have configured in my parent disks to host windows 2012 and some common utilities. Thereafter, with a few lines of PS script, I can create my diff disks for hosting SharePoint 2010, 2013, Office Web Apps etc .. .I want to have some robust back ups so I can regress after applying Service Packs and CUs .
Now given the fact the windows has to read from both the parent and diff disks, but write to the diff disks. I gather creating snapshots ( yet another disk created ) would really be bad for performance. Hence, what should I back up (see below) and is there a way to schedule this automatically; with or without stopping the VMs.
Incidentally, I had a peek at one of my VM folders and thought what the hell?
Daniel
I'm trying to install a Hyper-V Generation 2 guest OS Windows Server
2012 on a host OS Windows Server 2012 R2 RTM from the technet
subscription.
What I found interesting from the description of the Hyper-V generation 2 is "This virtual machine generation provides support for features such as secure boot, SCSI boot, and PXE boot using a standard network adapter. Guest operating systems must be running at least Windows Server 2012 or 64-bit versions of Windows 8"
However, when I tried to install a Windows Server 2012 VM, my keyboard is not working and if I want to copy something from the clipboard I got below error.
I tried install Windows Server 2012 R2 as well, all works great.
So i'm wondering is it a bug for Windows Server 2012 installation or the above support guest OS description need to change to Windows Server 2012 R2?
Ricky
I am running two servers with Microsoft Hyper-V Server 2012 (core) and using Hyper-V Replica between the two servers. Both ends have 30/5 Mbps connections (down/up).
There is only one VM on each server that is being replicated. Most of the changes/data is accumulated on ServerA.
I initially setup ServerA and ServerB side by side for their first replication and then moved ServerB to the remote location.
Recently there was a lot of data added to ServerA (around 300-400 GB). Over the past several weeks more data has been added. In checking the Replica stats it showed a peak of around 980 GB.
Around September 9th the replication stopped while ServerB showed the VM was "Modifying Resource." This continued for several hours.
Afterwards the replication resumed. (I don't recall if I had to manually restart the replication or if it resumed on its own.) But I still had several hundred GB to replicate.
Last night the replication was around 95%. A few hours later I assume it must have hit 100% as when I checked I found the replication had stopped and ServerB was again showing the ServerA VM as "Modifying Resource."
I thought this should have meant the full 980 GB of data had finally been replicated. When the replication resumed it was as if a "snapshot" had completed. But in checking the stats I still saw the following.
Size of data yet to be replicated:616.64 GB
Last synchronized at:09/09/2013 10:22:01 PM
So after running constantly since at least September 8th the amount of replication completed was only about 1/3 of the amount of data. And the date only moved up by 6 days.
I haven't added any data to ServerA recently, but anticipate in the next week or two I will. At this rate I'm wondering if ServerB will ever catch up and the ServerA VM be fully replicated?
When the replication resumed last night it showed 2% completed. After running overnight (about 6 hours) it only shows 4% completed.
When making some estimates about performance I thought 5 Mbps upload speed should work for the relatively small amount of data I need to replicate. But now I'm not sure it is going to be adequate to keep up?
Especially after seeing that once the replication reaches 100% that doesn't seem to mean that all the pending data I see is going to have been sent. Only a "snapshot" is likely to have completed?
Has anyone else done similar testing and have you got any insight you can share about these topics?
Theokrat
Hi,
I cannot add any new hardware to running VM in SCVMM.
I want to add SCSI controller and then HD.
Also in integration services I uncheck Time Synchronization OK. No error message. When check it again, the check mark is there.
Why it happens?
Thx.
As per windows server 2012 R2. The reference image if often seen,
Does Microsoft recommends to store VMs on SMB 3.0 share instead of block level CSV ?
Windows Storage spaces can use shared SAS to emulate SAN features, but then does it then requires that storage be utilized as SMB 3.0 Share only ?
I believe the virtual disk created using storage spaces can be utilized as CSV also.
Shahid Roofi
Hi All,
I have a two node Hyper-V Failover cluster that is taking advantage of SMB 3.0 for the VHD storage of our VM's. Just the other day the VM's in the failover cluster will randomly go into a "stopped" state and then reboot. They do not fail over, they just reboot on the node that they are on at a random time during the day.
Another thing I had noticed is that when I purposely fail a VM over via Live Migration or Quick Migration it does successfully complete, BUT after this happens the permissions that have been granted for both nodes in the cluster lose "Full Control"
access to the SMB file share with the VMs stored on it. So the only way to rectify the issue is to manually grant Full control back to each node to the "C:\VM Storage" share.
I also tried creating another VM fresh with nothing special and it still will randomly reboot without warning. I can't seem to dig up anything in the event logs either. The only thing that is recurring is something along the lines of "Unable to access specified share C:\VM Storage\ etc..." Which I seem to have found the root cause of that issue. Would anyone have any insight on this issue, or possibly have experienced something similar before?
Thanks!
