dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
5477
share rss forum feed


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

1 edit
reply to Davesnothere

OFF-PEAK Speed Tests, April 13th

 
Note that in the START Off-Peak tests, just above, that Toronto and Kitchener did poorly, where Eastern cities (closer to me), and oddly also Hamilton (which is not near), all did well in Downstream Speed.

This makes me suspect a problem with strange routing or peering WITHIN THE COGECO AGGREGATED NETWORK, as these were ALL off-peak tests, and I repeated the poor cities after the good ones, to triple check.

OTOH, in the COGECO Off-Peak tests, further above, The Eastern cities were also better, but by a much smaller margin.

From this set of results, I would also be curious as to whether START's TorIX link is behaving as it should be.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

1 edit

 
I got to thinking further, that it could also be some oddball routing, peering, and loading, of some of the volunteer servers which we/they are using thru this test site.

Inconclusive then.

AND of course, there's the matter of the D2 modems contibuting to node channel congestion (a big reason for the design of D3 modems) - that's where a big chunk of MY money would go !



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

1 edit

 
Here are a few more of START, with the Motorola D2 modem :
START - D2 - 16/1 - 11:00 AM EDT
[Test Location in Eastern Ontario]
NOTE : ALL Locations were tested at least 2 times (especially if a result seemed too good or too bad to be true) and often with browser page refresh between each test of the same city.

I DID get a single read of 16 Downstream on Kanata this time, but it was not consistent.

 
Kitchener


Kanata (Ottawa)


Montreal (FibreNoire)


Toronto


Hamilton



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

REAL-TIME Speed Tests, April 13th

 
On my temp 16/1 profile and Motorola D2 modem on START's Cogeco feed, this morning I pulled in some torrents at a sustained 1.8 MB/s - I did not time them though.

Pretty good for that profile, and with no 'Speed-Boost', I'd say.



Gone
Premium
join:2011-01-24
Fort Erie, ON
kudos:4

Yeah I still have no issues pulling 1.8-1.9MB/s on usenet.


Phorkster
Premium
join:2004-06-27
Windsor, ON
kudos:1
reply to Davesnothere

Re: Cogeco to START - Changeover Logistics

I'm not a happy camper right now. My Start service was supposed to be on today, but it isn't.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

2 edits
reply to Davesnothere

Speed Tests, April 13th, Near Peak

 
START - D2 - 16/1 - 10:20 PM EDT
[Test Location in Eastern Ontario]
NOTE : ALL Locations were tested at least 2 times.
 
Kitchener


Hamilton


Kanata (near Ottawa)


Montreal (FibreNoire)


Toronto



Gone
Premium
join:2011-01-24
Fort Erie, ON
kudos:4




Things are fine here. Are you sure you aren't suffering from some sort of D2 node-level congestion?

Also, when I run the speedtest over wifi on my laptop, I only pull 6. My hard-wired desktop is doing fine and pulling the speeds above.


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

2 edits
reply to Davesnothere

 
COGECO - D3 - 30/2 - 10:40 PM EDT
[Test Location in Eastern Ontario]
NOTE : ALL Locations were tested at least 2 times.
 
Kitchener


Hamilton


Kanata (near Ottawa)


Montreal (FibreNoire)


Toronto



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits
reply to Gone

said by Gone:

Things are fine here. Are you sure you aren't suffering from some sort of D2 node-level congestion?

Also, when I run the speedtest over wifi on my laptop, I only pull 6. My hard-wired desktop is doing fine and pulling the speeds above.

 
I think that several factors are at play, one of them the D2 Local Node 'stuff-up', plus my own physical location vis-a-vis Start's & Cogeco's networks and peerings, as well as other things.

Also note my followup Cogeco tests of the same 5 servers.

The comparisons for tonight follow pretty much the same patterns as the ones which I did and posted for just after 11 PM last night.

It would interest me very much to have somebody near Toronto to post the same detailed tests of START as I did, using the D2 modem and the same 5 servers, to compare all of the stats for each.

Nudge, Hint, Mister Gone, Anybody else....


(BTW, ALL of my tests were done with the same PC, wired to my WRT54GL router, wired to whichever modem.)


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

 
Is Last Parade up and running on START yet, with the fancy D3 Cisco beast ?



rocca
Start.ca
Premium
join:2008-11-16
London, ON
kudos:23
reply to Davesnothere

Keep in mind from you to Toronto is the same path. The difference is from our peering in Toronto to specific sites, and the capability of those sites.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits

said by rocca:

Keep in mind from you to Toronto is the same path.

The difference is from our peering in Toronto to specific sites, and the capability of those sites.

 
I was alluding to some of that just above.

I have not performed any traces so far, but have you noticed a few patterns in my 2 nights of tests ?

My best PING with START's hookup is to Montreal, and usually my best speed too if during evening - Kanata is next best, and Toronto ties with Kitchener for worst speed during evg. - Speeds on START are much more even during the daytime or any other off-peak period.

On Cogeco's hookup, my best PING is to Toronto, though SPEEDS are much more consistent from one city to another, and at all times of day or night.

The main question which comes to MY mind is why my PING on START is lowest to Montreal if my data has to first go to Toronto on Cogeco's AGG network, and then back to Montreal, but a Toronto test shows poorly and it's just 'across town' from wherever Cogeco peers, I would think.

It must be taking a short cut to Montreal somehow, or a wrong turn in and around Toronto....

So how would I find out the name or IP of the various speedtest.net servers, so to do ping trace routes for each ?


Gone
Premium
join:2011-01-24
Fort Erie, ON
kudos:4

Perhaps because the Toronto server does not prioritize response to ICMP requests and/or is under much heavier load?

The bolding isn't necessary. Seriously.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits

said by Gone:

Perhaps because....

 
The Toronto test server responds just fine when I'm on my Cogeco account and modem - in fact the lowest ping of them all.

Did I miss something obvious ?

BTW, I post so much stuff in rapid succession some days that occasionally I really DO want portions of it to stand out more than the rest, and often the other text formatting choices just don't seem to cut it.

OK, I moved one of the tags.

I wish that there was a 'Semi-Bold' for Internet.


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7
reply to Gone

said by Gone:

Perhaps because....

 
So how would I find out the name or IP of the various speedtest.net servers, so to do ping trace routes for each ?


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits

Speed Tests, START, April 14th, OFF-PEAK

 
START 16/1 account with D2 Motorola modem - 2:00 AM EDT
[Tested from Eastern Ontario]
NOTE : ALL Locations were tested at least 3 times.
 
Montreal (FibreNoire server)




Toronto



Moments before this, my Cogeco account and D3 modem tested at full speed (30/2) on both of these servers, with PINGs of around 18 for Toronto and 26 for Montreal (FibreNoire server), and that observation seems to be holding true no matter what time of day.

WEIRD ?


Last Parade

join:2002-10-07
Port Colborne, ON
reply to Davesnothere

Re: Speed Tests, April 13th, Near Peak

said by Davesnothere:

 
Is Last Parade up and running on START yet, with the fancy D3 Cisco beast ?

No, I can't wait for my 9 hour install window on the 18th though.

spikernum1

join:2010-03-03
Burlington, ON
reply to Davesnothere

Re: Cogeco to START - Changeover Logistics

how odd is this
michigan server:

vs.
toronto


the odd part is that it is consistent.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits

said by spikernum1:

how odd is this
michigan server vs. toronto

the odd part is that it is consistent.

 
So is the type of result which I posted just above yours.

And you appear to be in Burlington, which geographically is apparently the closest point to where START connects to the Cogeco AGG POI.

Hmmmm.... [Dave scratches head]

MY result using the START account is consistently favouring Montreal for lowest PING, regardless of time of day.

On one hand, this makes sense, since I am closer to Montreal, but OTOH, I was told above that my START account data all goes first to Toronto on Cogeco's network, where it peers to START's network, and so on....

The SPEED may vary due to time of day, but the PING relationship remains that way.

But on my Cogeco account, it consistently favours Toronto for PING, and the speed hardly EVER drops below spec, EVEN during peak period.


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

1 edit

HOW to Do It

 
The main question which comes to MY mind is why my PING on START is lowest to Montreal if my data has to first go to Toronto on Cogeco's AGG network, and then back to Montreal, but a Toronto test shows poorly and it's just 'across town' from wherever Cogeco peers, I would think.

It must be taking a short cut to Montreal somehow, or a wrong turn in and around Toronto....

So how would I find out the name or IP of the various speedtest.net servers, so to do ping trace routes for each ?

I will repeat this post periodically until someone offers a way, or tells me that there IS none.

Thanks.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

4 edits

11 AM Tests - START account

 
START 16/1 account with D2 Modem :
[Tested from Eastern Ontario]

(However, my Cogeco account and D3 modem continue to test at full speed (30/2) on both of these servers, with PINGs of around 18 for Toronto and 26 for Montreal (FibreNoire server), and that observation remains true no matter what time of day.)

Toronto :



Montreal (FibreNoire server) :



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

2 edits
reply to spikernum1

Re: Cogeco to START - Changeover Logistics

said by spikernum1:

michigan server:

 
Just for Kicks, here is my test for Mount Pleasant, Michigan (Winn Telecom server) (same city that you tried) :

Lower ping than Toronto, but higher than Montreal.


Gone
Premium
join:2011-01-24
Fort Erie, ON
kudos:4
reply to Davesnothere

Re: Speed Tests, April 13th, Near Peak

said by Davesnothere:

The Toronto test server responds just fine when I'm on my Cogeco account and modem - in fact the lowest ping of them all.
Did I miss something obvious ?

Yes. Cogeco has drastically different peering arrangements than Start. Cogeco doesn't peer at Torix, for example. Who knows what kind of route you're getting to get to that server, or what or what port you're using if it's multihomed. You're posting decent speed tests from other servers, so who the hell cares if but one server has lower performance than the other ones you're posting?

Seriously, when it comes to stuff like this there are so many variables that were this conversation in person it would be nothing more than an exercise in oxygen/CO2 exchange.


rocca
Start.ca
Premium
join:2008-11-16
London, ON
kudos:23
reply to Davesnothere

Re: HOW to Do It

said by Davesnothere:

 
The main question which comes to MY mind is why my PING on START is lowest to Montreal if my data has to first go to Toronto on Cogeco's AGG network, and then back to Montreal, but a Toronto test shows poorly and it's just 'across town' from wherever Cogeco peers, I would think.

Latency rarely has much to do with distance when distance is less than a few hundred miles.

Also FYI, Cogeco's aggregation network terminates in Burlington, not Toronto.


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

2 edits

said by rocca:

....Also FYI, Cogeco's aggregation network terminates in Burlington, not Toronto.

 
Understood - Though I regarded Burlington as 'across town' from Toronto when I said that.

BUT, when closer geography from test sites to my ISP and Cogeco's AGG POI seem to affects my ping AND speed negatively, even in off-peak periods, I feel that it matters and is worthy of investigation.

However, we'll know better on the SPEED part after the D3 modems are happening, so I'll try to be patient.

I just want to make sure that as much info is 'Out There' as possible, and these tests are an important part of it, IMNSHO.


Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7

 
So how would I find out the name or IP of the various speedtest.net servers, so to do ping trace routes for each ?

I will repeat this post periodically until someone offers a way, or tells me that there IS none.

I want to LEARN something from this, even if it turns out to be less important in the big scheme of things than I perceive it to be.

Thanks.



rocca
Start.ca
Premium
join:2008-11-16
London, ON
kudos:23
reply to Davesnothere

I don't think the 20ms latency issue will have any impact on what one test server says over the other.

As Gone mentioned, there are multiple paths, different arrangements, even paths between two directly connected routers can take varying physical paths through different physical switches and even different physical paths between cities. There are too many variables for me to be able to say with any authority why one server sitting somewhere on the internet is giving better performance than some other server sitting somewhere else on the internet. I know nothing about their connectivity, their load, their paths, etc.

Speedtests are a good baseline for 'is my connection giving me the speed I've ordered' and if one of them says yes, then you know you're local connection has the correct speed. They are also good for if you're seeing poor performance surfing the web then you can use it as a quick check to see if you're generally having problems. However, if you repeatedly speedtest over and over and over using different sites you will undoubtedly find that sometimes you get different results - especially on cable.

In any case, the new 20ms latency that popped up last week is being investigated but as everyone seems to be reporting good speeds and no impact, 20ms (we are talking milliseconds here, ie 1/50th of a second) really shouldn't affect day-to-day in the meantime.



Davesnothere
No-BHELL-ity DOES have its Advantages
Premium
join:2009-06-15
START Today!
kudos:7
reply to Davesnothere

 
OK, I tried this while on my START account :

tracert to the root domain of one provider of a test site

FibreNoire in Montreal :

Tracing route to fibrenoire.net [208.84.104.2]
over a maximum of 30 hops:

1 - 3 ms 1 ms 1 ms 192.168.67.1
2 - 8 ms 8 ms 20 ms 10.205.36.1
3 - 43 ms 43 ms 43 ms 64.140.112.154
4 - 48 ms 44 ms 42 ms 64.140.112.153
5 - 44 ms 44 ms 42 ms bvi1.asr01.tor151f.fibrenoire.ca [206.108.34.141]
6 - 51 ms 56 ms 50 ms te4-3.crt01.mtlcx03.fibrenoire.ca [68.67.63.197]
7 - 78 ms 50 ms 52 ms gi6-9.core02.mtl-1250.fibrenoire.ca [68.67.63.254]
8 - 51 ms 48 ms 49 ms srv01.mtl-1250.fibrenoire.ca [208.84.104.2]

Trace complete.



rocca
Start.ca
Premium
join:2008-11-16
London, ON
kudos:23
reply to Davesnothere

said by Davesnothere:

 
So how would I find out the name or IP of the various speedtest.net servers, so to do ping trace routes for each ?
I will repeat this post periodically until someone offers a way, or tells me that there IS none.

You could install a packet sniffer on your computer and see where the packets are going.