VMware Cloud Community
SurfControl
Enthusiast
Enthusiast
Jump to solution

When one leg of the network resumes connectivity, causing host ping timed-out...

We have vSwitch0 configured with two port groups:

Management Network and vMotion

There are two pNICS (vmnic0 & vmnic1) associated with vSwitch0 and the setting for nic teaming is all defaults.

(vmnic0 – Active/vmnic1 –Active) /Load Balancing – port-based

The Problem:

I have the continuous Ping pinging the host.

if I UN-PLUG one physical nic's network cable, I see may be 1 or 2 drops from the ping, b/c the other pNIC picks up the connection so no problems here.

but if I PLUG this network cable BACK-IN, after 2 or 3 seconds, I'll see about 7-request-timed-out from the ping, then it resumes the ping after.

that's about 21 seconds and that’s long enough to trigger a host isolation(?)

I understand that I can go into the HA Advanced Options to change the das.failureInterval to a bigger value,

but my question is why, when the connectivity resumes, it would cause long delays for the host to become pingable...

is this long delay a normal behavior?

I did try to set nic teaming FAILBACK to NO, but that didn’t seem to help…

the cluster I have is a two node esxi 4.0u1 w/ vCenter 4.0u1

Thanks,

Reply
0 Kudos
1 Solution

Accepted Solutions
Exwork
Enthusiast
Enthusiast
Jump to solution

This may be caused by spanning-tree relearning your network connectivity from bringing up the second NIC. I have seen similar issues.

View solution in original post

Reply
0 Kudos
3 Replies
lnairn
VMware Employee
VMware Employee
Jump to solution

Hi,

are you sure that the best Load Balance algorithm for you infraestructure is port based?

i had some issues regarding to load balance algorithm.

Reply
0 Kudos
Exwork
Enthusiast
Enthusiast
Jump to solution

This may be caused by spanning-tree relearning your network connectivity from bringing up the second NIC. I have seen similar issues.

Reply
0 Kudos
SurfControl
Enthusiast
Enthusiast
Jump to solution

thanks for the info...I found a kb article "[STP may cause temporary loss of network connectivity when a failover or failback event occurs|http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1003804]" which does sound like where the problem is but i have to check with the network folks to confirm the settings on the switch...

Reply
0 Kudos