Hello, gents.
I am out of ideas.
I have very weird issue with networking on freshly installed w2012r2 VM.
And I've spent several hours trying to determine on which level there is an issue and I am still have no full clarity.
The issue is simple - freshly installed VM cannot see network and ping gateway.
I've tried it on 3 ESXi 5.5 (1474528) hosts and on those hosts there are other VMs w2012r2 and w2k8r2 are running and working fine with network.
On this VM it just show Unindetified network and that's it.
I've tried different things: change portgroups, re-create vNICs (both as E1000E and VMXNET3), install/uninstall VMware Tools, upgrade vHW, migrate to other hosts, reinstall from different ISO images, checked networking configuration on host level, tried different IPs in guest OS, disable IPv6, set netowork speed and duplex, recreate routes in Windows and other stuff ...
VM able to ping it's own IP but nothing on the network.
Something i am missing ... but nothing comes to my mind.
What am I missing? Any ideas?
Thank you in advance!
Thanks everybody, VMware support has assisted to find the issue.
One of the vmnics is not working, we were able to create separate portgroup and add only one vmnic to it.
With one vmnicX network communication works with another vmnicY it fails.
Thanks everybody for your prompt responses and willingness to assist, this wasn't an easy one!
Try to change the machine's IP address.
I've tried to use different IP, from w2012r2 VM in the same cluster, which works fine with networking and it didn't solve the issue.
Disabled Windows Firewall and still no result.
Take 2 VMs that are on the same host / same port group. One works (A), the one that doesn't (B).
- On A :
- On B :
(Have you checked there's no duplicate of this IP somewhere on your network?)
Can you provide a screenshot of where it displays "unidentified network".
Did you deploy the VM from a template or manually ?
Can you verify you are on the correct subnet? Maybe with a SS of both the working server and the one that is not working.
Disabled Windows Firewall and still no result.
Network config is identical, except for IP.
172.20.1.43 and 172.20.1.216.
mask 255.255.255.0
GW 172.20.1.1
Duplicate network would show message in ipconfig /all and it shows (Preferred).
Tried to use working VM IP on not working VM with no luck, so it's not IP related.
Validated networking settings several times, and re-deployed new VMs, setting up new parameters again and again, same issue.
It's French Windows Server, so it looks like this:
VMs were deployed manually from 2 different ISOs French and English, same issues.
Tried this:
Switched to private, didn't change anything.
How did you create the other servers (that work) in the first place by the way?
Can you ping 172.20.1.43 to/from 172.20.1.216?
Is there some kind of radius mac based authentication on your network or any other "third party" security policy?
I'm running out of ideas but I'm thinking more and more that it's something silly, always is.
It's not me who installed them, they were deployed before. But when I though it was ESXi host related issue and I need to plan updating ESXi, I saw these.
No, from new VMs I am able to ping only their own IPs, nothing else.
No, as far as I am aware.
Opened case with VMware support, but I am afraid they will send me to MS support.
This is certainly a strange one and I am sure we have missed something. If you have logged a ticket with VMware, then they will at least confirm it is not a VMware issue. However, I would like to request you to review troubleshooting steps with the help of the article below :-
Troubleshooting virtual machine network connection issues (1003893) | VMware KB
If it is on the same port group as another one that works in the same subnet then it's very unlikely to be hypervisor related.
Maybe you can attach the VMX of a vm that works and the one that doesn't to see if we find something but again I think your ESXi/VM is fine.
One last thing to try would be to try using the mac address of a VM that works.
Same port group as others.
My team has found weird workaround, but I still have no valid explanation why it works ...
VM start working normally if you add second vNIC adapter with different IP.
That's insane ... I hope VMware support can add some light ...
How are you getting your network to the ESX hosts. Access ports, trunks, etc. Are you using vSwitch tagging?
Try this; create a private virtual switch and portgroup with no uplinks and attach this VM and another using the same subnet and having them ping each other. This could rule out some type of ARP issue on the switching gear. It would also rule out the vSwitch where the other VMs are.
Also, just want to confirm that we are not doing anything with VDS right?