Experiment: pfSense OpenVPN vs Wireguard “RoadWarrier” SpeedTest - Methods and Results

kevdog

Introduction

I've documented an experiment I ran testing pfSense's OpenVPN setup vs Wireguard in a typical roadwarrior setup.

An iphone8 using Verizon's LTE network was used as the client using either OpenVPN Connect or Wireguard apps as downloaded through the Apple AppStore

SpeedTests were measured using Ookla (https://speedtest.net) against same server.

Parameters tested were:

Control: SpeedTest of iPhone8 over Verizon LTE directly

OpenVPN: SpeedTest of iPhone8 tunneled through home pfSense via OpenVPN

Wireguard: SpeedTest of iPhone8 tunneled through home pfSense via WireGuard

OpenVPN was configured to use TLS1.3

Wireguard configured with basic parameters.

Methods and Results (Documented here):
https://forums.lawrencesystems.com/t/experiment-pfsense-openvpn-vs-wireguard-roadwarrier-speedtest/11638

Conclusions:

Control:
iphone8 over Verizon LTE -
Download 77.5 Mbps, Upload 12.1 Mbps

OpenVPN:
iphone8 tunneled via OpenVPN / Verizon LTE through home pfSense -
Download 78.8 Mbps, Upload 9.85 Mbps

WireGuard:
iphone8 tunneled via WireGuard / Verizon LTE through home pfsense -
Download 18.7 Mbps, Upload 4.88 Mbps

Questions:

Is my test setup accurate? Did I misconfigure my setup to draw incorrect conclusions?
Why does Wireguard Tunnel perform so much poorer than OpenVPN? There is virtually no difference between OpenVPN vs control. With Wireguard tunnel, Download speed was about 4 times slower and Upload Speed about 3 times slower.
Does Wireguard need additional configuration parameters (MTU adjustments??)