Update – April 3/2015

I’m not so sure about this anymore. Since writing this article I had a Rogers tech over and he replaced the cable in the back yard. My Rogers connection is now a lot faster. > 30Mb/sec download. My A/B tests between VPN and no VPN are roughly the same since. I’ve seen some links to this post out there so I should mention my latest findings.

After reading Michael Geist’s article about a senior VP at Rogers calling on the Canadian Government to shutdown VPNs it occurred to me that I have been experiencing chronically crappy network connectivity when working from home (from which I am always connected over the easyDNS corporate VPN) and thought maybe if this sentiment exists in the upper echelons of Rogers management, then maybe it has trickled down to a policy of traffic shaping along those lines. In other words, maybe my crapola connection over VPN is because Rogers is throttling VPN traffic.

My setup at home is that I use Rogers cable as my main internet connection and our backup is Teksavvy DSL.

So I ran the following speedtests:

1. Rogers
2. Rogers over VPN
3. Teksavvy
4. Teksavvy over VPN

What I was interested in was how much of a slowdown using my company VPN affected each connection. If Rogers wasn’t throttling VPN it should be roughly the same amount of slowdown. I used speedtest.net and in all four tests drew the same testing node from Beanfield, right here in Toronto, so it made for a nice test environment.

Here are the results:

Rogers bare:

Rogers over VPN:

Teksavvy bare:

Teksavvy over VPN

It looks like Rogers loses slightly over 5Mbps or > 50% of it’s download speed over VPN, while Teksavvy loses much less, just 1 Mbps or 5% in percentage terms. Teksavvy outperforms Rogers by far in either scenario. Not sure why I didn’t test this before, because the obvious play here is to switch my main connection at the house over to Teksavvy and use Rogers as the backup connection, not the other way around.