I was having this problem on an Ubuntu x64 host. I got fed up with no support, so i switched everything to KVM. I started experiencing the same issue with KVM. After alot of research and pulling out hair, i figured out it was the sky2 driver installed by default in ubuntu. I updated the network driver with the one on Marvell's official website and no problems since. Network is slow, but it doesn't quit. Not positive if this will solve the vmware problem since i'm on kvm now, but it's worth a shot.

Also after seeing the amount of support we received in this thread, i think i'm going to stick with KVM.

1) Isn't this thread about vmware networking bridges dying? Did i post in this in the wrong thread? My network bridge was dying, then i switched to KVM.

2) VMWare doesn't support Linux? Are you joking? Why can i download the linux vmware tarballs off their website then?

3) No S-H-I-T dumbass, neither are most of the solutions in this thread. I posted my solution in hopes that it might help someone else.

Was i really complaining? I'm so glad i don't have to deal with VMWare support and dumbasses like yourself anymore.

Why do you post here if you don't have anything of value to add?

http://kvm.qumranet.com/kvmwiki
http://en.wikipedia.org/wiki/Kernel-based_Virtual_Machine

yes, same abbreviation

it`s more a matter of "when will they start listening and make some qualified person to join this thread to get into the issue"

there is silence for months now from vmware , and that`s the main problem.
they don`t have interest in fixing this one.

fixing it is a differnt story, because i think this thread may not be about a single problem.
maybe different issues/bugs meet here.... (related to host os, related to user error.....or perhaps related to a vmware bug...)

I installed several waves of Microsoft updates to our server last night (SBS 2003). All four of the VMs stopped responding on the network, but could talk to each other. They all used bridged networking and static IPs, running Ubuntu 7.10. A quick Google landed me here. After reading this thread and trying various solutions I found that nothing seemed to work.

Then I read devzero's post about the ARP issue (http://communities.vmware.com/message/920916#920916). Running "vnetsniffer /e vmnet0" showed me the VMs sending out ARP requests but never getting anything back. So I was able to confirm his findings with my system. But what to do about it? Cycling the bridge device didn't help, nor did changing the MAC addresses of the VMs.

Just for a lark, I shut down all the VMs and (at random,) added a second bridged NIC to one of them. After firing it up - everything worked. I powered it off, removed the NIC, and tried again - still working. I then fired up all the VMs, and they are all working. Since I've been under pressure to get these back up I haven't tried a full reboot of the host server.

I hope this helps some other people out. I guess something in the VMNet arp table/cache got flushed or reset when the new NIC was installed - whatever it was I'm glad for a sudden flash of "whatthehelli'lltryit-ness"

• Jon

hi,

I've just stumbled into this discussion, since we are having the exact same problem (still), but only on a linux host and guest system.
see my details: http://communities.vmware.com/thread/141700

What driver is everyone using for their VM's. I've been using the e1000 driver. I've been working with support on this issue and now have become their test subject since they can't reproduce in their lab. Is anyone running the vmxnet driver or has anyone tried?

Basically these are the changes VMware wants me to make:

ethernet0.allow64bitvmxnet = "TRUE"
ethernet0.virtualDev = "vmxnet"

Thoughts?

I attempted changing from the e1000 to the vmxnet driver for my VMs. The problem I had with this was that the Windows 2003 x64 guests were not able to recognize the device and install drivers for it. I tried reinstalling the vmware tools in hopes that it would install the right driver for the vmxnet interfaces, but this did not work. I also mounted the vmware tools iso and copied the vmxnet drivers over to the windows server and tried to manually install them for the vmxnet devices. This did not work for me.

I am at a loss at where to go next with this issue and have tried all of the relevant suggestions in this thread. We purchased a support contract with vmware and they have not been able to help me with this yet either. After being bounced from technician to technician I finally got a call from the support manager saying that they would like me to run the vm-support.vbs script while the VM is experiencing the drop in network connectivity. Has anyone else with a support contract been asked to provide the same information? and if so what was vmware's response/finding? They say that they cannot recreate the issue in their test environment.

Another question: Has anyone that has been experiencing this problem tried moving their VMs over to an ESX server? Or to another virtualization platform? If so what platform have you moved to?

Good to know, thank you. We have been working with VMsupport since the beginning of February this year trying to correct this problem. I am not optimistic that vmware support is going to be able to fix this anytime soon and so we are considering changing virtualization platforms.

How long have you been on the ESX platform now without any drops?