V2T migration failed on host migration phase with error messages: management kernel interface <VMK_#> shares MAC with pnic <vmnic_#>

Products

• VMware NSX Networking

Issue/Introduction

Symptoms:

• V2T migration failed with the below error:

Execution of v2t script failed for host <HOSTNAME> [Errors: LACP is disabled on DVSwitch. NSX-t VSIP agent already present. RTM enabled: True /tmp/nsx/v2t_vnics_migration.json: { "migrate_to_ndvs": "true", "host_switches": [ { "host_switch_name": "<SWITCH_NAME>",

...
vmknics on the host: {'vmk2': 'vDS-Storage-A', 'vmk3': 'vDS-Storage-B', 'vmk0': 'vDS-Compute', 'vmk1': 'vDS-Compute', 'vmk4': 'vDS-Compute'}

...

Failed; error messages:\nmanagement kernel interface vmk0 shares MAC with pnic vmnic0, but the pnic is not in the active list of default uplink teaming of NVDS [<UUID>]. \n" } opsAgent CLI nsxa/reload_options output [Reloaded nsxa options from file ] ]

...

Environment

VMware NSX-T Data Center

Cause

• The teaming policy of DVS port group to which vmk0 is connected was found to have one link active & the other one is standby.
• On the NSX-T side, the default teaming policy as failover order and the active uplink on the default Teaming policy & the Management teaming policies active/standby uplink which was created manually after import migration stage was not matching.

Resolution

This issue is resolved in VMware NSX-T Data Center 3.2.2 and VMware NSX 4.1.0, available at VMware downloads.

Workaround:

• If the V2T migration has not reached the host migration step yet, rollback to the start of the migration and follow the steps below:

1. Remove all uplink teaming names from transport zones and VLAN segments.
2. In vCenter UI, find the DVS port group to which the vmk0 connects, then change its uplink teaming policy to 'loadbalance srcid' with both uplinks in the active list.
3. Start the migration from the beginning and let V2T take full control of the uplink teaming policies: V2T will create the uplink profiles then add the uplink teaming names in transport zones and segments.

 

• If the V2T migration already failed with the above error for a host, then follow the below steps:

1. Identify the management kernel interface, as listed in the error message, which is vmk0 in the above error message. Next identify the vmk (e.g. vmk0) inside the "vmknics on the host" part of the error message and get the VDS name e.g. 'vDS-Compute' in the above error message.
2. Go to NSX UI and find the uplink profiles used by the VDS in System -> Fabric -> Profiles -> Uplink Profiles, there should be two profiles named like "<VDS-name>-UplinkProfile-VXLAN_0" and "<VDS-name>-UplinkProfile-_0", edit them to change the teaming policies to “load balancing source” from “failover order” and add both uplinks in the active list.
3. The transport node of the host should be updated automatically, wait for its state to become success.
4. Go to the V2T host migration page and retry the host migration.