Test Fortify with the VPN off and on to find the bottleneck

To test a Fortify router properly, run the same speed test over Ethernet, Wi-Fi near the router, and Wi-Fi in your problem room—first without the router VPN, then with it enabled. That gives you six comparable test conditions rather than one number mixing coverage, device limits, internet service and VPN routing.

Fortify is the ExpressVPN–GL.iNet co-branded GL-MT6000 Wi-Fi 6 router. Its advertised “up to 900 Mbps” WireGuard speed is not a guaranteed internet speed: ExpressVPN says it measures the hardware’s VPN capability in client mode under controlled local-network conditions. An actual ExpressVPN connection also depends on the server, ISP, device and network conditions. See Fortify’s specifications and speed footnote.

Set up a fair comparison

Use one laptop for all six conditions, with an Ethernet adapter for the wired runs. Keep these conditions consistent:

• Internet connection: Record your plan’s download and upload speeds. Note whether Fortify connects upstream by Ethernet, repeater mode or tethering; these are different test conditions. Fortify supports all three in its setup guide.
• Device and link: Record the laptop’s Wi-Fi standard, band and channel width where available. Check the negotiated Ethernet link speed. Testing beyond 1 Gbps requires suitably fast connections throughout the wired path, not just a fast router port.
• Background traffic: Pause large downloads, cloud backups and streaming for the baseline tests. Disconnect any VPN app on the laptop so the VPN-on result measures the router tunnel alone.
• Test service: Use the same service and, where selectable, the same test server. If automatic server selection changes after enabling the VPN, record that change.
• Repetition: Run each condition three times, one test at a time, and compare the median—the middle result—not the best run. If results vary widely, retain the range and repeat later.

Device and adapter capabilities can limit results below both the router and internet-plan speeds, as Google Fiber’s speed-test guidance explains.

For throughput comparisons, use a test that reports download and upload. For connection quality, add a test reporting idle latency, loaded latency and jitter, such as Cloudflare’s. Keep those results separate: Cloudflare says its test measures connection quality rather than trying to saturate the link for maximum throughput.

Turn Fortify’s VPN off, then on

These dashboard instructions apply to co-branded Fortify, not every GL-MT6000: ExpressVPN says other Flint 2 GL-MT6000 models do not include this integration.

At the default admin address, 192.168.8.1, open VPN → VPN Client Profile → Go to ExpressVPN Dashboard. ExpressVPN’s Fortify instructions say the tunnel’s power button turns it off or on; the status changes between Not Connected and Connected. Other active tunnels may still route your traffic, so check every tunnel that could match the test device. Fortify tunnel controls.

For the VPN-off baseline, confirm the laptop can access the internet outside the VPN. Fortify’s Enhanced Kill Switch blocks traffic that does not match a VPN tunnel; Allow Non-VPN Traffic permits that traffic through the regular WAN connection. Each tunnel also has its own kill switch for connection failures. If your settings prevent a baseline, do not treat the blocked connection as a speed result. Make temporary policy changes only if you understand their scope: changing global settings can affect other devices, not just the laptop. VPN-off testing sends traffic outside the VPN. Restore your protection settings when the comparison is finished. GL.iNet’s Fortify policy guide.

For the VPN-on runs:

1. Activate one tunnel using a chosen nearby VPN location.
2. Make sure its device rules include the laptop and select All Targets for this comparison. If using Specified Devices, check the laptop’s identity on both Ethernet and Wi-Fi: GL.iNet identifies selected devices by MAC address, so verify the rules after switching connections. Fortify device and destination rules.
3. Verify that the laptop’s public IP changes on each connection type, and record the connected VPN location after every run.

The built-in ExpressVPN integration uses WireGuard, according to the Fortify-specific support guide. Do not assume protocol options advertised for desktop apps also apply to this integration.

Run the six tests

For wired tests, disable the laptop’s Wi-Fi. For wireless tests, unplug Ethernet and verify the band; use 5 GHz near the router if supported, keeping the same band in the problem room where possible. If the band changes, record it rather than treating distance as the only changed variable.

Test position	VPN off	VPN on
Ethernet to Fortify	Baseline through the router and ISP	Adds the router’s VPN processing and VPN route
Wi-Fi, roughly 1–2 metres away in the same room	Checks near-router wireless performance	Checks whether Wi-Fi or the VPN path limits speed
Wi-Fi, usual problem location	Checks the effect of coverage	Measures performance where you actually use the VPN

Record download and upload in Mbps, plus idle latency, loaded latency and jitter in milliseconds. Also note time, band, VPN location and test server. Our three-test Wi-Fi guide explains the underlying comparison.

Read the pattern before changing anything

• Wired VPN-off is already slow: Check the plan, Ethernet link, upstream connection and test server first. This does not establish a VPN problem—or prove the ISP is at fault.
• Wired is fast, but nearby Wi-Fi is slow: Investigate the laptop’s wireless capability, band and radio conditions. Compare a second device before replacing the router; see why a phone can be slower than a laptop.
• Nearby Wi-Fi is fast, but the problem room is slow: Coverage is the stronger suspect. Try repositioning the router and retest before upgrading the internet plan.
• VPN-off is fast, but wired VPN-on drops sharply: Test another nearby VPN location while keeping other variables fixed. The added VPN path or processing is implicated, but one result cannot identify the router hardware as the cause.

For a numerical comparison, calculate VPN retention = VPN-on download ÷ VPN-off download × 100 using the medians from the same test position and service. A hypothetical 480 Mbps result against a 600 Mbps baseline retains 80%; it is not an independent Fortify benchmark or a pass/fail threshold.

Finally, check responsiveness. Loaded latency measures round-trip delay while traffic is flowing; jitter measures variation between delay samples. A hypothetical rise from 20 ms idle to 150 ms under load adds 130 ms of delay that the download headline hides. Cloudflare’s methodology explains these metrics.

For gaming, compare the game’s own ping to the same region with the VPN off and on. A speed-test server is not the game server, and a VPN changes the route rather than guaranteeing lower ping—a limitation acknowledged in ExpressVPN’s gaming guidance.