vmkernel.log on ESXi host actively reports asynchronous hot reset events initiated by the bnxtnet driver:
vmkernel: bnxtnet: bnxtnet_handle_hot_reset:#####: [vmnic0] Hot reset async evt: evt_data1 = 0####, min wait time = 1300 ms, max wait time = 6200 ms vmkernel:
bnxtnet: bnxtnet_handle_hot_reset:#####: [vmnic1] Hot reset async evt: evt_data1 = 0#####, min wait time = 1300 ms, max wait time = 6200 msVMware ESXi 8.x
The bnxtnet driver is triggering an error recovery process (bnxtnet_handle_hot_reset) in response to a firmware crash. One possible reason behind firmware crash is incompatibility between the physical NIC firmware version and the NIC driver version.
Identify the current bnxtnet driver version and the applied physical NIC firmware version on the affected ESXi host.
Validate the driver and firmware combination against the Broadcom Compatibility Guide. This can be verified at below link:
Broadcom Compatibility Guide for IO Devices
If the current firmware version is not validated for the installed driver version, update the NIC firmware to a supported release.
Japanese version of this KB : https://knowledge.broadcom.com/external/article/436408