Network failures in ESXi configured with host profile
search cancel

Network failures in ESXi configured with host profile

book

Article ID: 341204

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

Update the host profile from the reference host and remove certain host-specific parameters to resolve the network failures on the ESXi.


Symptoms:
Inbound and outbound network fails on virtual machines hosted an ESXi host configured using a reference host profile.

Environment

VMware NSX for vSphere 6.0.x
VMware NSX for vSphere 6.1.x
VMware NSX for vSphere 6.2.x

Cause

This issue occurs when RMQ settings are specific to host NSX setting. When using host profiles, RMQ settings must be disabled, so that host profile does not overwrite host specific settings.

Resolution

To resolve this issue, update the host profile from the reference host and remove certain host-specific parameters, after preparing hosts for NSX and VXLAN configuration.
 
To update the host profile from the reference host and remove certain host-specific parameters:
  1. Right-click the host profile and click Enable/Disable Profile Configuration.
  2. Go to Networking configuration > Host Virtual NIC.
  3. Deselect the VXLAN vmkernel interface to ensure that NSX Manager always creates the VTEP on boot.

    Note: if your reference profile contain multiple VTEPs, ensure that all VTEPs are removed.

  4. Go to Networking configuration > NetStack Instance > VXLAN.
  5. Deselect the ipRouteConfig option.
  6. Under Advanced configuration option, deselect all UserVars.Rmq* variables:

    # grep Rmq /etc/vmware/esx.conf
    /advUserOptions/options[0030]/name = "RmqClientPeerName"
    /advUserOptions/options[0031]/name = "RmqHostId"
    /advUserOptions/options[0036]/name = "RmqClientResponseQueue"
    /advUserOptions/options[0037]/name = "RmqClientExchange"
Note: These variables are host-specific.
 
Refresh the host profile and verify that NSX parameters are ignored when the host profile is applied.
 
Notes:
 
  • Apply the default value to all netstack instances except for the default TCP/IP stack when setting the hostname. To do so, VMware recommends you to navigate to Networking configuration > NetStack Instance > VXLAN > DNS Configuration > Hostname, select Obtain hostname from DCHP.

  • When configuring a host for VXLAN in the NSX Manager UI, ensure that the configured LACP setting matches the setting on the virtual switch. If the settings differ, in the NSX Manager logs, you see entries similar to:
<YYYY-MM-DD>T<time>.650 GMT WARN http-443-exec-22 DefaultExceptionLogger:35 - The following exception occurred during request processing by the BlazeDS MessageBroker and will be serialized back to the client: </time>
flex.messaging.MessageException: Invalid LACP version selected. Distributed Virtual Switch vDS-xx is configured with different LACP version.
 
Unpreparing a Host for VXLAN

To unprepare a host for VXLAN, you must remove the VXLAN network configuration parameters, such as IP route config and VTEP interfaces, and three NSX VIBs from the image profile and create a new image profile. If VIBs are not removed from the image profile, the host show as Not Ready in the Host Preparation tab after reboot. Clicking on the error next to the Not Ready link displays the reason for the error.