VM disconnects ethernet from router

[thestewman]

I recently added a Windows 7 Pro VM to my Unraid 6.1.6 system.

The system has run flawlessly for 2 years with no problems.

I found after the VM installation that if I used the VM either with RDP or VNC sometime later the system appeared to be hung.

All leds etc are normal but the GUI or web page will not load. I can connect to the system via IPMI as it has a separate ethernet connection to the Supermicro MB. I cannot run a CMD line task.

A reboot via the case power controls or via the IPMI starts the system. This of course causes Unraid to start a new parity check.

 

I found out today if I unplug and then reconnect the ethernet cable the system is running OK

 

I found this entry in my system log which shows the disabled state and when I unplugged and reconnected the cable.

 

I have attached the System Log and Diagnostics

Dec  9 16:37:41 Tower unmenu[17631]: cat: /sys/block/loo/stat: No such file or directory

Dec  9 21:25:51 Tower kernel: e1000e: eth0 NIC Link is Down

Dec  9 21:25:51 Tower kernel: br0: port 1(eth0) entered disabled state

Dec  9 21:26:04 Tower kernel: e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx

Dec  9 21:26:04 Tower kernel: br0: port 1(eth0) entered listening state

Dec  9 21:26:04 Tower kernel: br0: port 1(eth0) entered listening state

Dec  9 21:26:19 Tower kernel: br0: port 1(eth0) entered learning state

Dec  9 21:26:34 Tower kernel: br0: topology change detected, propagating

Dec  9 21:26:34 Tower kernel: br0: port 1(eth0) entered forwarding state

Dec  9 22:09:52 Tower emhttp: cmd: /usr/local/emhttp/plugins/dynamix/scripts/tail_log syslog

Dec  9 22:11:15 Tower unmenu[17631]: cat: /sys/block/loo/stat: No such file or directory

 

I would apprecaite if someone can figure out what is happening and direct me in how to correct this ethernet disabled state or disconnect.

syslog-2015-12-09.zip

tower-diagnostics-20151209-2216.zip

[fireplex]

Sounds like same issue as http://lime-technology.com/forum/index.php?topic=39890.0

[thestewman]

fireplex

Thanks for the link. Yes,  I had read that posting previously prior to my own experience and after spending an hour using the forum search tool I could not find the original post. Guess my search query was to vague or something.

Anyway I used some suggestions from other Unraid users and I think from JonP.

The disconnect has not returned and it has been several days

Here is what I did

I turned Off DHCP in my router for the unraid IP address and made my connection to the server Static

In Setup >Network here are my settings

MAC address: XX:XX:XX:XX:XX:XX

Enable bonding: Yes

Bonding mode:  Active Backup (1)

Setup bridge:    Yes

Bridge name:      br0

Bridge enable STP: Yes

Bridge forward delay: 0

Obtain IP address automatically: No

IP address: 192.XX.X.XX

Network mask: 255.255.255.0

Default gateway: 192.168.1.1

Obtain DNS server address automatically: No

DNS server 1:  8.8.8.8

DNS server 2:  8.8.4.4

DNS server 3:

 

 

 

[fireplex]

I was already running with a static setup anyway and tried the STP and bonding options but made no difference fort me.

[accoleon]

I had a similar issue, watching videos (or anything network intensive) on either of my VMs would cause the unraid network to disappear (the VMs can ping each other and Tower, but none of them can access the internet), while the rest of the devices (outside of the unraid box) can access the internet and network fine.

 

I have an Airport Extreme, and turning off ipv6 on the router resolved all my network issues. I guess ipv6 isn't enabled in the unraid kernel, and once the router starts sending ipv6 frames, the unraid server started ignoring them and became nonresponsive network-wise until an ifconfig eth0 down/up.

 

Hope this helps.

[fonts]

i had the exact same issue.

after researching for a while i stumbled upon a thread that said to go back to a previous version of the vm guest drivers

these are my results

virtio-win-0.1.102 - no issues (this is the previous stable version)

virtio-win-0.1.110 - vm's drop out at random times. ( this is the latest driver last time i checked )

so roll back your vm drivers version and you should be all sweet.

 

[stanger89]

i had the exact same issue.

after researching for a while i stumbled upon a thread that said to go back to a previous version of the vm guest drivers

these are my results

virtio-win-0.1.102 - no issues (this is the previous stable version)

virtio-win-0.1.110 - vm's drop out at random times. ( this is the latest driver last time i checked )

so roll back your vm drivers version and you should be all sweet.

 

Thanks for this!!!!!!

 

I've been chasing this for ages, and I thought I had those drivers installed, but I just forcefully "updated" my Windows 7 VM's Virtio drivers to the 102 version and it appears to have fixed my issue.

[Ned]

I have the exact same issue with my Windows 10 VM.  My windows 7 VM is just fine on virtio-win-0.1.96 (actually when I tried to upgrade the ethernet driver through Windows device manager in the VM it caused a blue screen of death I could not recover from, thankfully I had a backup).

 

I updated the drivers for ethernet, scsi, etc. in device manager to the virtio-0.1.102 drivers but it did not help!  Still after a few hours unraid loses network connectivity and I need to unplug and reconnect the ethernet cable to get it back.

 

Not sure I want to try the 110 driver since it has not been reported to help.

 

What else can be causing this??  Are these drivers compatible with Windows 10?  Is unraid simply not ready to properly support Windows 10 VMs?

[Ned]

Just updated to virtio 113 drivers which now have Win10 drivers and it appears to have fixed the problem!