NIC index on Server 2008 core keeps changing and, as a result, virtual clients cannot find the gateway.
- I have installed W2K8 core on a system with 2 NICs. I have configured the Hyper-V role and updated the Core server. I have been able to configure several virtual guests and they worked fine for a while (a few hours). But then they lost the ability to ping their gatewway and so lost their internet connnectivity. I am just configuring the virtual network adapter as an external adapter. No other virtual network configurations.
What I have found so far is that the physical NIC is changing its index. I disabled the 2nd NIC through the Core command line and rebooted the Hyper-V Core server. At that point, none of the VM's would start and I had to delete and recreate them. The virtual guests network adapters now seem stable, however I am wondering just what did happen? Has anyone else experienced this?- Type modifiéVincent HuMSFT, Modérateurmardi 27 octobre 2009 09:00
Réponses
Hi,
If you are a new to Hyper-V, I recommend that you install a full installation of Windows Server 2008, and then enable Hyper-V.
By the way, the information you provide is not enough to perform the further research, please provide more information about your environment.
Vincent Hu
- Marqué comme réponseVincent HuMSFT, Modérateurmercredi 28 octobre 2009 07:22
Toutes les réponses
Hi,
If you are a new to Hyper-V, I recommend that you install a full installation of Windows Server 2008, and then enable Hyper-V.
By the way, the information you provide is not enough to perform the further research, please provide more information about your environment.
Vincent Hu
- Marqué comme réponseVincent HuMSFT, Modérateurmercredi 28 octobre 2009 07:22
- Hi Vincent,
Actually, I've worked with both the GUI version and Core. That is to say that I installed 2 partitions on the boot drive and have configured it to dual boot. The default brings up Server 2008 Enterprise 64 bit (not R2) GUI and the 2nd partition boots to Hyper-V Core (same server version). The boot drive is 80GB and the data drives are 2 - 1TB drives configured as a 2TB striped drive that has been hardware configured by the on-board RAID provided by the ASUS P6T Deluxe V2 motherboard. I store my VM's and VHD's on the striped drive. The physical NIC's are Marvell Yukon 88E8056 PCI-E Gigabit Ethernet Controllers. I have disabled controller #1 (under Core) and no longer have the Index change problem as noted earlier.
What happens now is that the VM's will, after a period of time usually less than 1 day, end up loosing there ability to ping the defeault gateway (192.168.1.1). What is very strange is that while the VM's loose that capability, the Core Server does not. I can continue to ping the gateway and also internet addresses just fine from the Core. Since my understanding is that the Core (the host server) is also a virtual server, it does not make sense that it would behave differently than guest VM's.
I run Hyper-V manager from a Vista PC and the Core Server, Vista PC and all other VM's are configured as members of the workgroup WORKGOUP. I login to each with the same administrative username and password. I have updated the Core with all required Microsoft updates via the VB script found on the MSDN site. I don't know what other information you may need that would prove helpful but I appreciate your input and if there other specifics you would want, please let me know.
Thanks, Olympic Bill Hi,
I think I get your concern now. If I misunderstand your concern, please feel free to let me know.
Now the issue you encountered was that the VMs can’t ping the default gateway after a period of time.
In order to perform the further research on the issue, please help me to collect and confirm the following information:
1. You have two physical NICs on the computer, now you have disabled one, it seems that you created one External Virtual Network bound to the rest one physical NIC, and manage the Server Core computer through the virtual network adapter for host communication.
2. What’s the version of your problematic VMs?
3. Did you encountered this issue with all the VMs?
4. Did you find any errors inside the VM in Event Viewer?
By the way, please perform the following suggestions to see whether you get the same issue:
1. Boot the problematic VMs into safe mode.
2. Remove the current network adapter of the VM, create a new network adapter, you may try to use legacy network adapter if you are using synthetic network adapter.
3. Disable TCP offloading on both the physical NIC and the virtual network adapters.
4. If possible, please insert one external PCI NIC to the computer and then create a External Virtual Network, connect the problematic VMs to it.
Vincent Hu
