O Canada! > Canadian > TekSavvy > [Cable] Slow speed on Ajax POI during evenings?

[Cable] Slow speed on Ajax POI during evenings?

Hello,

Try some trace routes during the time where speeds are subpar and we can see where the issue is occurring.

If you wish you can post in the Direct forum as well.

Thanks,
Paul

reply to destrukt
Hello,

There have been no reports of issues on the AJAX POI. It is currently running under 65% capacity.

Martin

reply to TSI Paul
Just now, tested at 7mpbs, here's the trace.

Tracing route to google.com [74.125.226.5]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms homerouter [192.168.1.1]
2 11 ms 9 ms 8 ms 10.125.226.1
3 21 ms 46 ms 18 ms 69.63.242.233
4 19 ms 8 ms 10 ms ajax1.cable.teksavvy.com [24.246.55.70]
5 20 ms 20 ms 11 ms 24-52-255-69.cable.teksavvy.com [24.52.255.69]
6 47 ms 13 ms 21 ms 72.14.212.134
7 19 ms 11 ms 13 ms 216.239.47.114
8 16 ms 12 ms 12 ms 72.14.233.142
9 13 ms 12 ms 16 ms yyz06s05-in-f5.1e100.net [74.125.226.5]

Trace complete.

reply to TSI Martin
It's happening right now. Just did a speedtest at 8.8mbps with no other utilization on my line.

Here's a traceroute to google as in my last note.

Any idea? I'm seeing more posts here in the forum about slow cable speeds in the evenings.

Tracing route to google.com [74.125.226.73]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms homerouter [192.168.1.1]
2 33 ms 6 ms 8 ms 10.125.226.1
3 15 ms 15 ms 15 ms 69.63.242.233
4 102 ms 102 ms 108 ms 24-52-255-70.cable.teksavvy.com [24.52.255.70]
5 19 ms 16 ms 12 ms 24-52-255-69.cable.teksavvy.com [24.52.255.69]
6 13 ms 25 ms 26 ms 72.14.212.134
7 13 ms 38 ms 15 ms 216.239.47.114
8 13 ms 36 ms 17 ms 64.233.175.98
9 17 ms 18 ms 15 ms yyz06s07-in-f9.1e100.net [74.125.226.73]

Trace complete

reply to destrukt
Using a docsis 2 or 3 modem? Maybe it's local node congestion? It's already been stated that they're not over capacity at the POI level.

Docsis 3 - the Thomson 475.

Just speedtested again now and it's at 15mbps, which is better than earlier, but still not full (normal) speed.

Trace:

Tracing route to google.com [74.125.226.78]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms homerouter [192.168.1.1]
2 12 ms 10 ms 27 ms 10.125.226.1
3 15 ms 13 ms 11 ms 69.63.242.233
4 9 ms 9 ms 8 ms ajax1.cable.teksavvy.com [24.246.55.70]
5 11 ms 11 ms 18 ms ajax1.cable.teksavvy.com [24.246.55.65]
6 13 ms 17 ms 17 ms gw-google.torontointernetxchange.net [206.108.34.6]
7 20 ms 16 ms 19 ms 216.239.47.114
8 23 ms 13 ms 13 ms 64.233.175.98
9 11 ms 14 ms 11 ms yyz06s07-in-f14.1e100.net [74.125.226.78]

Trace complete.

So am I up the creek here?

Nothing has changed in my internal network, and the speed returns to normal outside of the evening hours.

I use VOIP for my house phone and to guarantee available bw for that purpose, I use QoS. This condition is rendering my QoS settings ineffective due to the of the high variability of top end speed throughout the day.

TSI: Is there any additional diagnostic data that I can provide from my end that would help come to root cause on this issue?

You could always post in the »TekSavvy Direct forum. We can have you run a few tests.

Ok, reposted all of this there.

Speeds great for me - earlier this morning
»www.speedtest.net/result/2216804260.png
during peak times (upload is bad because I was uploading something)
»www.speedtest.net/result/2217360269.png

reply to destrukt
Seconded, getting 28 no problem right now.
It's not the POI. Its your local node and/or wiring/hardware.

reply to destrukt
It has to be the local node, signal levels to the modem, or the modem itself.

If you have a router, rule that out by going direct from your computer to the modem.

reply to destrukt
Just food for thought, but I pretty much put up with this issue too on and off for the last 3 months from the Ajax POI. Since my local node was upgraded to include both upstream & downstream channel bonding (8x4 vs 4x1), the problem seems to have disappeared.

Similarly manifested - OK speeds during non-peak hours, then degradation down to ~6-10 Mbps during peak. Same package, same POI. Even after link upgrade @ the POI, problem was not resolved. Modem direct to computer didn't change a thing - so the issue most likely had something to do with node over-subscription.

Before: