Why does Wi-Fi connection speed decrease substantially when 1.1.1.1 WARP enabled?

VPN
1

Can anyone explain why internet connection speed decreases 50% to 60% when I enable 1.1.1.1 WARP?

I’m running Wi-Fi 6e on a laptop with Windows 11 to a TP-Link DECO 6e mesh router. When I engage 1.1.1.1 WARP speeds drop dramatically 50%-60% as measured with SpeedTest.net. When I connect by ethernet there is only about a 10% degradation.

When I disable 1.1.1.1, speeds return to expected levels and are comparable for Wi-Fi and Ethernet. So, it is definitely the Cloudflare app in Wi-Fi mode.

Anyone have any clues why speeds drop so much through W-Fi? If so, is there anything I can do about it?

Any insights will be greatly appreciated.

Thanks,

PFB

2

Do provide numbers from speed.cloudflare.com with and without WARP.

3

Are you testing on the same server with and without WARP ?

Otherwise the only issue I can think of is an MTU issue maybe

4

I would guess MTU or TCP flow control. Perhaps a buffer problem? Does your Ethernet go into the same router/ ap?

5

1.1.1.1 make some of the website (secure) not able to load. Very bad

6

Yah, I think the real question is: Which bunch of lunatics got loose from the asylum and hired into the CloudFlare marketing department?

Claims about VPN making internet faster are simply … fantasy.

7

Using speed.cloudflare.com. Speeds are download. San Antonio to Dallas. Average of 2 attempts.

With Warp: Wi-Fi - 217 Mps Ethernet - 507 Mbps

Without Warp : Wi-Fi - 595 Mbps Ethernet - 616 Mbps

Thanks

8

Thanks for the suggestion. Same gear on my end and I tested to the same server locations on the receiving end.

Regarding MTU, since the connection speed is roughly equivalent for Ethernet and Wi-Fi without WARP, and the degradation also occurs when WARP is enabled for Ethernet or Wi-Fi, wouldn’t that mean that packet size is not a factor?

Or, is MTU configured differently in WARP for Ethernet than for Wi-Fi? How would I be able to determine this? I couldn’t find it in the WARP Advanced settings.

9

Thanks! Yes, both Wi-Fi and Ethernet are flowing through the same router. I’m not a network pro so I will review TCP flow control and buffer possibilities.

I thought I had addressed the MTU possibility earlier. I could not find any settings in the WARP app that could change MTU, and, since Wi-Fi without WARP is running essentially the same as Ethernet without WARP it didn’t seem that the Wi-Fi packet settings are the issue. I could certainly be wrong, and I will dig deeper.

What I don’t know is whether or not the router is treating Ethernet traffic different than Wi-Fi traffic because they are assigned different LAN addresses. I don’t see why the router would manage Ethernet differently than Wi-Fi since I didn’t change any settings. I am very possibly wrong with this assumption.

Thanks again for the ideas to investigate.

10

VPN adds overhead. Maybe the Cloudflare name servers are faster … but hard to imagine that the Cloudflare network is faster than the internet.

11

Thanks! Indeed, WARP is geared towards a Web browsing usage, where 200+ Mbps is much more than enough.

More than 595 Mbps would be a 1 Gbp VPN and even in the more costly providers, will also cap at 100-200 Mbps max.

12

Thinking about it … MTU is likely. 802.11 (better known as WiFi) has a MTU of 2304 bytes. Ethernet has the concept of “jumbo frames”, where the MTU can be set up to 9000 bytes. Packets are getting wrapped. So if you can’t support large frames on WiFi it is doubling the packets. Vs on Ethernet just adding the ~20% overhead for putting the packet in another packet.

13

Thanks for the input. You’re absolutely right, I’m happy with 200+ speeds sitting in my living room. But what about where I need the product most, like on a public or hotel Wi-Fi?

Besides, since this degradation is only happening on Wi-Fi with WARP, the only unexplained variable is how WARP is working with/managing the Wi-Fi connection. This appears to be is an issue with the application.

I’m a Cloudflare stockholder. The reason I invested in the company is because I believe in the company …its products, its operations, its strategy and its management. The reason I use WARP is to see how well it works. When I encounter an apparent problem with a company’s products, it concerns me.

As another commenter has pointed out, this problem runs counter to Cloudflare’s public statements and the image the company is trying to convey to the world.

If an amateur like me can perceive a problem with a Cloudflare product, what is a network security manager going to think?

14

Thanks for the tip on MTU. It’s always a good day when I learn something new.

I optimized MTU but am getting the same result… connection speed decreases 50% - 60% when WARP is enabled vs turned off.

I don’t think it’s a router or hardware issue because with WARP turned off 802.11ax is delivering the same speed as Ethernet. That means the radio connection is sound and throughput is not restricted through the router.

I can only conclude that WARP for Windows 11 is goofy. Additionally, when I run the same test, with and without WARP, on the same router from my iPhone Wi-Fi, the decrease is only about 10% or about what would be expected. But, on the Cloudflare website they say the software is different for iOS and Android.

I appreciate you taking the time to respond.