1

We've been having issues while setting up a vSphere cluster, using 7 ESXi hosts, and running about 74 guests that are amongst the mess.

Our issue is that when we reboot a guest, Windows or Linux, they come back and report an IP Address conflict.

We do not run DHCP, all hosts have statically assigned IP addresses. All network connectivity has been tested on all of the guests, but recently this issue has started plaguing us.

Can anyone shed some light on this issue?

squillman
  • 37,618
  • 10
  • 90
  • 145
Eirik W Toft
  • 11
  • 2
  • What MAC address is reported in the IP address conflict message? Is it the same as the guest that was rebooted? – EEAA Aug 13 '12 at 20:33
  • This can be caused by network switch, router connections, proxy-arp service etc – Andrew Smith Aug 13 '12 at 20:46
  • It is, actually. I made a point to write down the MAC address before rebooting the guest so I could verify that. – Eirik W Toft Aug 13 '12 at 22:19
  • But does this MAC address conflict with any other VM? If you clone a VM, VMWare will ask if you copied or moved it. The reason it asks, is so it knows whether to generate a new MAC address for the VM. If you copied a VM, but told VMWare you moved it, then you could have a MAC address conflict as the underlying cause – carpii Aug 13 '12 at 23:03
  • It does appear that a router was causing this issue. It's strange since that router isn't on the same network that it would have piced up and broadcasted these. Thanks for the assist everyone! – Eirik W Toft Aug 14 '12 at 00:04
  • That is true about the MAC address being cloned. One thing I learned early on is that cloning is a crummy, at best procedure - I've had much more luck making templates and deploying VMs from that. Aside from some small Linux trickery, I've had great success in deploying multiple VMs that way. – Eirik W Toft Aug 14 '12 at 00:05

1 Answers1

1

Have you cleared arp cache on the switch the ESX hosts are connected to? See, http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1008177

Chida
  • 2,471
  • 1
  • 16
  • 29
  • I did, but it only resolved the issue for a short time. It turned out there was a separate piece of network gear that I finally disconnected from the network and the problem went away. – Eirik W Toft Aug 14 '12 at 00:07