Long story short I have Rogers cable Internet 500 Ignite service in Toronto. I use it to connect to a corporate VPN (Cisco Anyconnect). With the VPN connected I noticed that there was high latency and slow throughput so I compared notes with the my team. We setup an internal iperf server and over the VPN the iperf test results were odd. An initial iperf test would deliver a consistent 20Mb/s, then subsequent tests in succession would return slower throughput. By the third consecutive test the results were terrible with an average speed of 5Mb/s for a 10 second test with iperf reporting 0Mb transferred for some of the test intervals. We then port forwarded iperf from the Internet, disconnected the remote endpoints from the VPN and got the same performance. I even ran the same iperf tests from an Android device on Rogers LTE with the Anyconnect VPN on and off using Hurricane Electric’s iperf tool and got the same results over wireless with a completely different endpoint! What gives? Meanwhile team members with Bell Internet were getting iperf test results of 120Mb/s on and off the VPN. I threw in the towel and had Bell install its Fibe 500 service yesterday and now my endpoint gets great iperf throughput results; 175Mb/s is the average. I realize that this is not a scientific result but I’m happy with it.
Hmm… not sure.
I would like to test here to compare, but our setup here is a little different.
We are using cisco anyconnect, but its set up as a split tunnel type setup.
Only data specifically going to something on the internal network, is going through the VPN connection, everything else is going out regular internet.
Doing a speed test, I am getting pretty much the same results between my personal machine and my work laptop. (different models, etc… so may be slight variances there), but that makes sense as the speed test will be going out the regular internet.
(and no cant compare with the work laptop with the VPN off, its a force VPN)
I’d like to share my experience.
I just have a 75 Mbps and most of the time it’s very slow, like 5 Mbps.
Once I was using IDM for downloading a big file and it was showing me around 220 KBps (220x8=1.76 Mbps) as soon as I started speedtest.net my speed went up to around 6MBps (6x8=48 Mbps).
First I didn’t believe it and tried it a couple more times, amazingly the same result every time.
I don’t know how they do it but as soon as Rogers detects a speed test site they increase the speed.
I don’t trust their speed test at all, and believe they are cheating.
I wouldn’t say that. This is just my experience, and my $0.02.
We don’t use a split tunnel. I realize that’s the preferred setup for the VPN. Still, it doesn’t explain the test results from iperf when the internal iperf server was port forwarded through the firewall avoiding the VPN altogether. Let me add that speedtests with fast.com, or speedtest.net would always return great results. It’s not hard for Rogers to QoS or rate shape in favour of those sites. I do that with voice traffic.
Any network provider can apply QoS / rate shaping policies that favour certain Internet services just as you experienced. This was the purpose of the Net Neutrality legal battle in the USA Net neutrality in the United States - Wikipedia to try an eliminate this practice. For example, Rogers Wireless may offer a deal on Apple Music that lets you stream unlimited songs over the wireless network without it going against your account’s data cap. Now, you may be an audiophile and prefer Tidal since it offers higher quality music than Apple Music but since Rogers doesn’t have a deal with Tidal your use of Tidal will come out of your data cap. Knowing that you begrudgingly opt for Apple Music and Rogers wins since it gets a kickback from Apple.
Well, I think it’s a bit different than what I said. I think when they see that we are checking the speed they raise it ( to fool us ) and very slow at other times.
I would. Rogers ducking sucks. Customer for 20 years. Would never go back unless forced by location. Customer service is shit. Product is shit. Prices are shit. Shit shit shit
Actually, what I wrote is exactly what you mentioned. Rogers or any network provider can use QoS / rate shaping policies to favour any website. In your example you refer to speed test sites which could include fast.com, or speedtest.net. The network provider allows for unfettered network bandwidth to those sites, but then rate shapes down access to websites that it does not favour such as Bittorrent sites, VPNs, etc. I brought in Net Neutrality from the USA to show what was lost because of it failing to be applied legally.