Hyper v server
I have a dell server (power edge r520) running Windows server 2012 with hyper v
The server runs 5 vms
I have two nice in the server - 1 nic is assigned to my main virtual network which is used by 4 of he 5 vms
The other is used by the 5th VM which I have on a different subnet.
I have been having an issue for a couple of months where 1/2 of the network seemed to be going down... I have a backup server onsite and I could still access that ... The quick fix was to simply reboot the single 24 port switch in the main comms room. The site is in the hospitality business and is open 24 x 7 which hampered troubleshooting.. We had replaced the switch but the problem persisted - the network would work fine for about 2 weeks and then we would have to reboot the switch again.
Anyway I happened to be close by the last time and it turns out that it is only the hyperV host ( and its vms ) that are losing connection to ten network. I can ping the rest of the network...
What can be wrong with my host that a reboot of the switch it is icon extend to can bring it back on line? The moment the switch comes back on my host server replied again and simply resumes working.. No reboot of host or VM s is required!
A simple reboot of the switch
The server runs 5 vms
I have two nice in the server - 1 nic is assigned to my main virtual network which is used by 4 of he 5 vms
The other is used by the 5th VM which I have on a different subnet.
I have been having an issue for a couple of months where 1/2 of the network seemed to be going down... I have a backup server onsite and I could still access that ... The quick fix was to simply reboot the single 24 port switch in the main comms room. The site is in the hospitality business and is open 24 x 7 which hampered troubleshooting.. We had replaced the switch but the problem persisted - the network would work fine for about 2 weeks and then we would have to reboot the switch again.
Anyway I happened to be close by the last time and it turns out that it is only the hyperV host ( and its vms ) that are losing connection to ten network. I can ping the rest of the network...
What can be wrong with my host that a reboot of the switch it is icon extend to can bring it back on line? The moment the switch comes back on my host server replied again and simply resumes working.. No reboot of host or VM s is required!
A simple reboot of the switch
I'm assuming your VMs have unique, static IP addresses as well, correct? An IP conflict could cause similar issues. Also, check to make sure that you don't have DHCP enabled on more than one machine. I have seen "rogue" DHCP servers issue IPs in a range that is either the same as the host network, causing multiple IPs of the same type, or issues an address of an existing server, or usurps the gateway address causing traffic to stall.
If you have a managed switch, it's possible that STP is killing ports. When things are down, check to see if they are disabled or if anything shows up in the logs. Maybe drop the switches output to a syslog sever so you can correlate that data when it does go down.
How often does it happen?
If you have a managed switch, it's possible that STP is killing ports. When things are down, check to see if they are disabled or if anything shows up in the logs. Maybe drop the switches output to a syslog sever so you can correlate that data when it does go down.
How often does it happen?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
That seems to be it. I have updated the nic drivers and disabled VM queues.. All well so far
Please be aware that when the Broadcom NIC driver gets updated it may flip the Virtual Machine Queues setting in the driver for each port back to ENABLED.
did you update the lan drivers on the Dell R520 server to latest ?
try these ... all the best