VMware Cloud Community
Pvajdic
Enthusiast
Enthusiast

Losing network connectivity after host migration

Hi,

We have a rather big issue with VM live host migration. It happens very often that after live migration NIC adapter is disconnected and we can't enable it. I'm talking about "Connected" option under VM settings. I click on check box "Connected" apply settings and if I open Edit Settings window again it is unchecked. "Connect at power on" stays checked. Sometimes it will help if I delete NIC adapter and re add it again. We are using ESXi 5.1 build 1483097 Enterprise plus and VCenter server 5.1. This occurs on standard switch and on vDS. At our last migration deleting NIC didn't help, we had to migrate VM to another host in cluster (we have 3 hosts in cluster). Problem is not OS related as we are using Windows Server 2008 R2, Windows Server 2012, Windows Server 2003. On this cluster we have 58 VMs. Anyone had this problem?

Thank you.

Reply
0 Kudos
5 Replies
vfk
Expert
Expert

Does this happen for particular type of OS i.e. Windows Server 2008 R2, OR Windows Server 2012 OR Windows Server 2003?   What vnic are you using e1000 or vmxnet?  Does it happen every time a VM is vmotion-ed to another host for all the VMs?  Do you only have one cluster?

--- If you found this or any other answer helpful, please consider the use of the Helpful or Correct buttons to award points. vfk Systems Manager / Technical Architect VCP5-DCV, VCAP5-DCA, vExpert, ITILv3, CCNA, MCP
Reply
0 Kudos
schepp
Leadership
Leadership

Hi,

your ESXi build (5.1 Update 2) contains a known bug.

Apply the patch listed in VMware KB: Hosts with virtual machines using the E1000/E1000E adapter have networking issues after u...

or use the workaround mentioned there.

Regards

Tim

Pvajdic
Enthusiast
Enthusiast

Hi vfk,

It doesn't matter which OS is. I forgot to mention that this happens on E1000 vnics. We just started to replace Windows 2012 servers with vmxnet3 because they are the first ones that were losing connectivity every few days no matter on vmotion. Also this doesn't occur on every vmotion, it is random issue on random vms.

We have 10+ clusters with over 50 hosts and 500+ vms. All our ESXi servers are on same build version. This cluster is behind firewalls, proxy and what not...so there is a lot of "non-standard" traffic and it is the only cluster affected by this issue.

Regards.

Reply
0 Kudos
Pvajdic
Enthusiast
Enthusiast

Hi Tim,

It seems this is it!! As this is production cluster with very sensitive vms it will be complicated to patch them within few days. Some vms are using RDM for Windows cluster so there is no live migration option. I will try this patch on our test cluster and report back.

Regards,

Petar.

Reply
0 Kudos
vfk
Expert
Expert

Yup, classic e1000 issues, we have first observed this issue on ONLY Windows 2012, I have not seen it happen for other OSes, definitely get yourself patched.  As for the clustered VMs with RDMs, you will need to schedule a maintenance for this, there is no live migration because of the scsi bus sharing.

--- If you found this or any other answer helpful, please consider the use of the Helpful or Correct buttons to award points. vfk Systems Manager / Technical Architect VCP5-DCV, VCAP5-DCA, vExpert, ITILv3, CCNA, MCP
Reply
0 Kudos