VMware vSAN 8.x
"pNIC vSwitch Port Drop Rate" checks the vSwitch stats, which can include valid drops, for example if the physical switch sends a packet that can't be delivered (such as multicast or unknown unicast). If a vmnic is not being utilized, the drop percentage calculated from all traffic being forwarded to the vmnic may appear high.
When the vmnic is being utilized and more traffic is being received, the percentage "pNIC vSwitch Port Drop Rate" would drop near 0% even if the same number of packet drops are still occurring.
This is an expected behavior if there is no VM/vmk traffic on the vmnic in question.
As per vSAN Performance Graphs in the vSphere Web Client Part 1 :
This graph shows percentage of Physical NIC vSwitch Port Drop Rate.
| Metrics | Metrics Description | Metric ID in API | Corresponding Entity Type | Corresponding Real-time Entity type |
|---|---|---|---|---|
| Inbound Packet Drop Rate of pNic vSwitch Port | Percentage of Physical NIC Inbound vSwitch Port Drop Rate. | portRxDrops | vsan-pnic-net | hr-vsan-pnic-net |
| Outbound Packet Drop Rate of pNic vSwitch Port | Percentage of Physical NIC Outbound vSwitch Port Drop Rate. | portTxDrops | vsan-pnic-net | hr-vsan-pnic-net |
Q: Would setting the uplinks for the vSAN vmkernel adapter to Active/Standby help mitigate this alert?
A: Keeping the uplink in an Active/Standby or Active/Active configuration makes no difference. This is because the pNic error health check monitors all the uplinks configured for the vSAN vmkernel's port group.