Connection issue In NSX Overlay Cause by ESXI Physical uplinks Rx packet drops
Article ID: 442278
Updated On:
Products
Issue/Introduction
Multiple ESXi hosts are experiencing network packet loss and connectivity degradation. This manifests as partial packet loss for VMs residing on an NSX Overlay bridge profile, caused by continuous receive errors on physical uplinks (e.g., vmnic0 and vmnic4).
The hostd.log records continuous increments of errorsRx for the physical adapters:
Wa(164) Hostd: [Originator@6876 sub=Statssvc.StatsCollector] Error stats for pnic: vmnic0
Wa(164) Hostd: --> errorsRx: 803473899
The nicinfo output validates extreme error counts at the hardware level:
vmnic0: Total receive errors: 803473933
vmnic4: Total receive errors: 716667505Environment
VMware vSphere ESXi
VMware NSX
Cause
The packet loss is caused by faults at the physical network adapter hardware layer (Layer 1/Layer 2), preventing frames from successfully reaching the ESXi networking stack. The nicinfo statistics confirm that receive packets dropped by the hypervisor are 0, while total receive errors logged by the NIC hardware are extremely high, pointing to a hardware, firmware, transceiver, cabling, or physical switch port issue.
Resolution
Engage the server hardware and physical network vendors to perform hardware diagnostics on the affected physical NICs, transceivers, cabling, and upstream switch ports.
Apply any necessary firmware updates identified by the hardware vendor to resolve the hardware-layer fault.
Additional Information
For further verification of supported driver and firmware combinations, consult the VMware Compatibility Guide (VCG).
Feedback
