NSX Datapath core dump generated on Bare Metal Edge with Mellanox NIC.

Products

• VMware NSX

Issue/Introduction

• Datapath does not become operational, so there is no network connectivity for the data plane.
• Mellanox NIC is used. 
• Log lines similar to the below are encountered in /var/log/syslog

  NSX 29588 FABRIC [nsx@6876 comp="nsx-edge" subcomp="datapathd" s2comp="intel-rte" level="WARN"] mlx5_net: Unable to recognize master/representors on the multiple IB devices.
  NSX 29588 FABRIC [nsx@6876 comp="nsx-edge" subcomp="datapathd" s2comp="intel-rte" level="WARN"] mlx5_common: Failed to load driver mlx5_eth
  NSX 29588 FABRIC [nsx@6876 comp="nsx-edge" subcomp="datapathd" s2comp="intel-rte" level="WARN"] EAL: Requested device 0000:17:00.0 cannot be used
  datapathd 29588 intel-rte [FATAL] PANIC in fpn_app_init():

• Core dumps will be generated with entries similar to the below will be observed in var/log/syslog:
  

  NSX 30821 - [nsx@6876 comp="nsx-edge" subcomp="node-mgmt" username="root" level="WARNING"] Core file generated: /var/log/core/core.datapathd.###.gz

Note: The preceding log excerpts are only examples. Date, time, and environmental variables may vary depending on your environment.

Environment

VMware NSX

Cause

This scenario is triggered due to an unknown issue with the Mellanox NIC that causes a process panic. 

Resolution

This is a condition that may occur in a VMware NSX environment.

 

Workaround

A full power cycle of the Bare Metal Edge should resolve the issue. A soft restart may not clear this issue.