asked on
Intermittent network connectivity issues on Hyper-V VM - Adapter paused
I have a Win Server 2016 Hyper-V VM that intermittently gets an Event 10 where the message says "Miniport NIC 'Microsoft Hyper-V Network Adapter #2' paused" Any direction on how to resolve this?
Someone recently posted an identical issue, so I wonder if there isn't something larger at play here.
https://www.experts-exchange.com/questions/29245910/Hyper-V-Network-Dropouts-on-Broadcom-Adapter.html
ASKER
Here is the driver information. I'm not sure what exact model it is at this point.
How many ports in the server?
Are any of the ports teamed?
For Broadcom Gigabit make sure Virtual Machine Queues is set to DISABLED for _all_ Broadcom Gigabit Ports as this will cause problems with VM network connectivity.
I have two very thorough EE articles on all things Hyper-V:
Some Hyper-V Hardware and Software Best Practices
Practical Hyper-V Performance Expectations
ASKER
2 ports, None are teamed, the VMQ is set to disabled on all ports.
I'll check out the articles. Thanks!
This is likely your issue:
This issue is resolved in Broadcom driver b57nd60a.sys version 16.8 and newer. In March 2015, Broadcom published driver version 17.0 for download. In April 2015, HP published version 16.8 of the driver for their affected network adapters. Contact your server OEM if you need a driver that's specific to your server.
Please snip and post the VMQ setting referenced.
ASKER
ASKER
Andrew Porter, we do have the lastest drivers so that is not the issue here. See earlier post for screen shot of current driver.
Philip Elder, this also confirms the status of the VMQ settings
Following this as I have the exact same issue.
Thank you. The disabling of Virtual Machine Queues is key as doing so in the OS does not improve matters.
The next step is to remove the vSwitch from the VM, delete the vSwitch, create a new one exclusive to Hyper-V (not shared with the host OS), and bind the VM's vNIC to that newly created vSwitch.
How many ports in the server? If four, then LBFO teaming two for management and two for a dedicated pair using the HyperVPorts algorithm would be a better setup.
ASKER
We only have two ports so there are some limitations there. One was not being used so I did create a new switch and assigned it as a dedicated port to the VM that's having issues two days ago. However that didn't resolve the issue.
To add to this I also re created the vSwitch and this did not improve our situation either.
What is the server setup?
* CPU(S)
* Total Memory
* RAID Controller & RAID array/logical disk setup
What is the VM's setup?
* VM vCPU Count
* VM vRAM Assigned
* VM .VHDX File(s) assigned: Size, Percentage Free
Out of curiosity are you using Atera as your RMM agent?
ASKER
Curtis, It's interesting that you asked about Atera as yes, we are using their RMM agent.
Philip, Here are the system specs:
VM Host 2 processors, 16 Core per processor, 64 GB Ram, Perc H750, Raid 0, 2 partitions (OS, Data)
VM 10 processors, 32 GB Ram, 2 vhdx files,
OS vhdx, 200 GB, 76GB Free
Data vhdx, 1996 GB, 578 GB Free
Uninstall Atera that’s what I’ve just done, it appears to have stopped the dropouts so far. I think the Atera agent has been effectively been flooding the network and the adapter has been shutting down to recover and then turn back on.
You will notice a load of informational messages in event viewer before the Miniport shutdown referring to Atera installing packages.
Uninstall it and let me know what it does.
ASKER
Thanks Curtis! That's helpful.
Nice job, Curtis!
ASKER
Removing the Atera RMM worked for this server as well. Thanks for your suggestion Curtis!
What is the make/model of the NIC in the machine?