FreeGuard Testing Methodology
Last verified: 2026-04-15
This page describes how FreeGuard tests VPN performance, security, and reliability across all server regions. Our testing methodology is transparent and reproducible.
Tools We Use
iperf3
Measures raw TCP/UDP throughput between the test client and a controlled iperf3 server placed at each target location.
ping / mtr
Measures round-trip latency and hop-by-hop packet loss using ICMP echo requests over a 60-second window.
DNS Leak Test
This check sends an A query for freeguardvpn.com to each listed public DNS-over-HTTPS provider and reports only response metadata. It cannot identify your device's recursive DNS resolver or whether a request uses a VPN tunnel.
WebRTC leak checker (headless Chromium)
Automates a headless Chromium session to capture local and public ICE candidates and compares them against the expected VPN IP.
IPv6 Leak Test
This single dual-stack endpoint does not prove whether IPv6 is tunneled or blocked. A real check needs an owned IPv6-only probe on each network.
Wireshark / tshark
Captures live traffic during each test run to confirm all packets flow through the encrypted tunnel and none bypass it.
Test Environment
Hardware
Dedicated bare-metal test machines with 8-core CPUs, 32 GB RAM, and 1 Gbps NIC — ensuring the client is never the bottleneck.
Network Conditions
Tests are run across multiple ISPs per source location, covering fiber, cable, and 4G/LTE connections to capture real-world variance.
Geographic Distribution
Tests are executed from 10+ source countries across Asia, Europe, and North America to reflect global user conditions.
Test Protocol
Each server location follows this standardized step-by-step procedure:
Connect to target VPN server
Establish the tunnel using the FreeGuard client. Wait 10 seconds for routing tables to stabilize.
Verify tunnel IP
Confirm the public IP matches the expected VPN server egress address before any measurements begin.
Run speed test (iperf3)
Execute a 30-second bidirectional iperf3 test. Record download, upload, and jitter.
Measure latency (ping/mtr)
Send 100 ICMP packets to the target. Record min, avg, max RTT and packet loss percentage.
DNS Leak Test
To investigate a DNS leak, compare your OS, browser and VPN DNS settings with packet captures or resolver logs on real networks.
Run WebRTC leak test
Launch headless Chromium, collect ICE candidates, and assert no real IP is exposed.
IPv6 Leak Test
This single dual-stack endpoint does not prove whether IPv6 is tunneled or blocked. A real check needs an owned IPv6-only probe on each network.
Disconnect and reset
Tear down the tunnel, flush routing tables, and wait 30 seconds before the next run.
Repeat 3× and average
Perform the full sequence 3 times per session. Compute the mean for speed and latency; treat any single leak as a failure.
Test Schedule
Tests are conducted monthly across all server regions. Results are updated on corresponding tool, gaming, and streaming pages within 48 hours of completion.
Sample Size
Each server location is tested a minimum of 3 times from at least 2 different source locations. Final metrics are the arithmetic mean of all runs, with outliers (±2σ) discarded.
How to Reproduce Our Results
You can verify our results independently using the same open-source tools. Here is the minimal setup:
Last verified: 2026-04-15