Help: Network not working after attempt at adding cluster

Ok, so I am new to VCenter and recently was trying to add a cluster and my networking seems to have stopped working between my VMs, however I can access them individually from my laptop via RDP/SSH/etc.

Backstory with pertinent details: I had 3 servers that I ran ESXi 6.5 Free on individually, two that I used for work related VMs and one for personal. @ of the servers are off most of the time, one is on 24/7. They each had a VMNetwork added with all VMs connecting to and using the same network (VLANS are coming) I then decided to join VMUG so I could get VCenter to help better manage the servers from one screen. This setup fine, and the three servers were added and everything was good. Recently I wanted to start playing around with Ansible for provisioning VMs and that is when I realized I did not have them in a Cluster. I started looking at doing that, and was able to create a cluster, which the system pulled all VMs in from the one server that is on. After this, I was able to provision from ansible and able to access the machines.


However, recently I tried to access a couple machines that talk to eachother and I noticed that they no longer communicate with hostnames, etc. In looking at VCenter, it seems like I need to possibly setup a Distributed Switch and move the VMs to that, rather than the VMNetwork switch that existed on the individual ESXi hosts.


Does this sound correct? If so, is there a recommended tutorial for how to accomplish this?

View Reddit by jimmyfloyd182View Source

Related Articles


  1. Having just migrated from vSwitches to VDS, the first question I have for you is:

    How are the hosts connected to each other?

    You don’t *need* VDS for a simple three-host cluster to talk to each other, but there may be some security settings that VC applied when it joined your hosts to the cluster.

    Second question: what are your security policy settings set to? (i.e. promiscuous, forged transmits, and whatever the third one is)

    I noticed packet-loss and reachability issues because I had static redundant links and I think my switches were getting confused with both ports from the host plugged in simultaeneously. Moving to VDS gave me functioning LACP to my switches. But that’s the only reason I even needed VDS.


    If you do migrate to VDS, move only one port at a time. First move one uplink port, then migrate VM networking into the port group, then move the second uplink port. I found that it was trying to move both, and vCenter couldn’t see the ESXI host anymore, and everything would roll back.

  2. Could it be as simple as your different hosts have slightly different vSwitch names. (ie a dot or a dash or a space in the name) If they are different and a vm is migrated to another host with even slightly different vSwitch name the network will not connect.

Leave a Reply

Your email address will not be published. Required fields are marked *