I've been meaning to post this for a while now, but I wanted to wait and collect a bit more data.
On January 6th I moved from Shaw (BB100) to Telus (Internet 50, fiber connection). Many people make qualitative statements when they speak about the differences between ISPs. I wanted to take the opportunity to do a quantitative measurement between the two ISPs.
I should say that this data isn't applicable to everyone. Shaw performs differently in different areas, but I do think it makes sense to actually measure your ISP's performance like this. These measurements are certainly relevant to anyone who lives in my neighborhood.
Anyway, on to the data.
The first chart shows the amount of data that was being sent or received, averaged by the hour. It is measured using the SNMP counters in my router (Asus RT-N66U running Tomato). There was ~538GB downloaded and ~86GB of data uploaded during this period.
The second chart shows the quality of the Internet connection by measuring the time between my router and the "first hop" on the network, which is the network gateway. Every minute PRTG (network monitoring software) would send 20 "ping" packets to the network gateway and record the min/max/average latency for the round trip. If this "max" number gets higher than ~50ms, that's very bad. It means that your Internet connection won't work well for online gaming or VoIP applications like Skype. Since the data is averaged in 1 hour intervals, you're not seeing the outlier readings. That is, when the max ping over 100ms, it means that the the average is over 100ms for that reading over an hour. There were many situations where I was getting max pings over ~1200ms in that data.
![Image]()
![Image]()
Notice how the Telus "ping" doesn't spike, even when I download lots of data (see January 21 for example). With Shaw, if someone started watching Netflix, that was all the traffic necessary for me to get lagged out (i.e. disconnected) from games of Starcraft 2.
It's also interesting that on the Telus connection there is some variability of my ping, irrespective of how much bandwidth I'm using. Note how on January 11th, January 14th, and January 16th the max ping moved up to about ~30ms. The rest of the time it was at ~4ms to ~8ms. I'm not sure what the root cause of this is, but it appears to be independent of whatever I'm doing.
That big red "downtime" spike is my router was disconnected when moving from Shaw to Telus on January 6th. The rest of the time there weren't any outages. I found that both Telus and Shaw were good from this perspective.
My recommendation: if you're a techie and you're currently on a cable ISP (or any ISP, really), it's worthwhile to setup something like PRTG (they have a free 10-sensor version) and keep tabs on how well your ISP is serving you.
I'm very happy with the move from Shaw to Telus, and the above data shows why. Not to mention that we went from paying $90+/month to $80/month ($45/month for the first 6 months). My loyalty to Shaw was really misplaced while I was waiting for them to improve their infrastructure. It turns out, they were in no rush to upgrade the network in my neighborhood (i.e do a node split).
On January 6th I moved from Shaw (BB100) to Telus (Internet 50, fiber connection). Many people make qualitative statements when they speak about the differences between ISPs. I wanted to take the opportunity to do a quantitative measurement between the two ISPs.
I should say that this data isn't applicable to everyone. Shaw performs differently in different areas, but I do think it makes sense to actually measure your ISP's performance like this. These measurements are certainly relevant to anyone who lives in my neighborhood.
Anyway, on to the data.
The first chart shows the amount of data that was being sent or received, averaged by the hour. It is measured using the SNMP counters in my router (Asus RT-N66U running Tomato). There was ~538GB downloaded and ~86GB of data uploaded during this period.
The second chart shows the quality of the Internet connection by measuring the time between my router and the "first hop" on the network, which is the network gateway. Every minute PRTG (network monitoring software) would send 20 "ping" packets to the network gateway and record the min/max/average latency for the round trip. If this "max" number gets higher than ~50ms, that's very bad. It means that your Internet connection won't work well for online gaming or VoIP applications like Skype. Since the data is averaged in 1 hour intervals, you're not seeing the outlier readings. That is, when the max ping over 100ms, it means that the the average is over 100ms for that reading over an hour. There were many situations where I was getting max pings over ~1200ms in that data.
Notice how the Telus "ping" doesn't spike, even when I download lots of data (see January 21 for example). With Shaw, if someone started watching Netflix, that was all the traffic necessary for me to get lagged out (i.e. disconnected) from games of Starcraft 2.
It's also interesting that on the Telus connection there is some variability of my ping, irrespective of how much bandwidth I'm using. Note how on January 11th, January 14th, and January 16th the max ping moved up to about ~30ms. The rest of the time it was at ~4ms to ~8ms. I'm not sure what the root cause of this is, but it appears to be independent of whatever I'm doing.
That big red "downtime" spike is my router was disconnected when moving from Shaw to Telus on January 6th. The rest of the time there weren't any outages. I found that both Telus and Shaw were good from this perspective.
My recommendation: if you're a techie and you're currently on a cable ISP (or any ISP, really), it's worthwhile to setup something like PRTG (they have a free 10-sensor version) and keep tabs on how well your ISP is serving you.
I'm very happy with the move from Shaw to Telus, and the above data shows why. Not to mention that we went from paying $90+/month to $80/month ($45/month for the first 6 months). My loyalty to Shaw was really misplaced while I was waiting for them to improve their infrastructure. It turns out, they were in no rush to upgrade the network in my neighborhood (i.e do a node split).
