TekSavvy → slow 6M DSL Montreal (evening only?)

DSL Tx 800 Rx 6016 AccConn: bas3-sherbrooke40

Line stat:
SNR 11.5 Atten. 37

voila
just got home and run a test

Same thing in Montreal here (eastern part), I'm on 16/1 DSL , and it's been like that for a week, on evenings.

The same thing here - Greenfield Park Qc. 1.94 down 0.61 up 34ms ping

I'm having the same issues...terrible speeds at night.
Came here to see if something was wrong...it's clear that something is going on.

Teksavvy, can you please let us know what's up?

Thanks

I just got an interesting finding,

I swap to a Bell 2wire modem and now the result is
I don't know if this will be an one-time "accident" or what?

Totally horrible speeds here too, 1mbps evenings and night, resorting to 3G tethering.

Slow speeds in Pointe-Claire right now as well.
~3.2/0.6 30ms ping

Dropped from 5Mbps to 3Mbps (on 6M/800 sync) since my last post. Pingtest jitter on the other hand is back down to my usual ~1ms.

I'm in Longueuil (Montreal south shore)

You have any tools to see want your upstream SNR and attenuation is ? I think its the IKNS bug you guys are experiencing !

Symptoms of the IKNS bug:

"Ikanos IKNS "Remote" DSLAMS have several bugs that result in false Line Stats produced by the CPE modem. These bogus Line Stats are frequently misinterpreted by inexperienced DSL problem solvers, resulting in misdiagnosis of (non-)problems, unnecessary trouble tickets, unnecessary modem replacements, unnecessary line repair calls, unnecessary anguish, unnecessary blaming, unnecessary unhappiness, and unnecessary human conflict. If there is no (other) problem with the line, equipment, noise, software, or configuration, the IKNS Dysfunctions are merely a nuisance. But if there are problems, the IKNS falsifications can seriously impede problem discovery and resolution.

Bell Canada has reportedly been aware of these bugs since early 2008. Ikanos have produced several failed non-solution attempts. The problem is regarded as low priority by Bell, since it is seen as not actually affecting DS data rates, and perhaps since it affects predominantly customers of Wholesaler ISP's, especially P2P users, and users who are aware of what data rates they are actually getting.

The Lantern Data, (accessible by the ISP), shows the Line Stats from the DSLAM end. It is the only source of correct Line Data, since the IKNS DSLAM falsifies that Line Data that it sends to the CPE modem!

. . .

The known bugs of IKNS DSLAMs show the following characteristic symptoms:

1. US SNRM is permanently falsely reported as 6dB.

2. US Attenuation is falsely reported as higher than DS. (Unless Attenuation is very low, DS Attenuation should be between 1.5 and 1.8 times the US Attenuation.)

3. US Sync Rate is lower than the Profile 800kbit/s, usually 512 to 768. (SpeedTouch modems tend to be immune from this IKNS bug).

4. Pathmode is permanently reported as "Interleaved", even though it is almost always "Fast Path". It may actually have been erroneously set to "Interleaved" by a technician who was fooled by the IKNS bogus reports of low US SNRM and high US Attenuation.

5. Increased latency, consistent with Interleaved, may result, depending on modem type. 2Wire modems tend to be particularly affected by this IKNS bug."

I am back again, and swap back with thomson tg784 and the speed drop back to 1M again.

So.....EdT, is this symptoms also helps in further determining the problem? thx

Terrible speeds downtown; getting 10 meg down on a 50 meg bundle, but it's not stable, it's super bursty. Tried reconnecting a bunch of times, even saw a result of 3 megs.

Considering all the speed issues people are having recently, even in Ontario, it seems like it might be a TSI issue.

Here are some details from both modems

Bell 2wire

DSL Line (Wire Pair): Line 1 (inner pair)
Protocol:G.DMT Annex A
Downstream Rate:6016 kbps
Upstream Rate:512 kbps
Channel:Interleaved
Current Noise Margin:24.0 dB (Downstream) 6.0 dB (Upstream)
Current Attenuation:13.1 dB (Downstream) 4.5 dB (Upstream)
Current Output Power:9.8 dBm (Downstream) 12.3 dBm (Upstream)
DSLAM Vendor Information: Country: {0xB5} Vendor: {IKNS} Specific: {0x01}
PVC Info:0/35

Thomson TG784

Uptime:0 days, 0:02:11
DSL Type: G.992.1 annex A
Channel mode: Interleave
Maximum Bandwidth (Up/Down) [kbps/kbps]: 1.056 / 11.008
Bandwidth (Up/Down) [kbps/kbps]: 512 / 6.016
Data Transferred (Sent/Received) [kB/kB]:29,00 / 69,00
Output Power (Up/Down) [dBm]: 12,0 / 9,5
Line Attenuation (Up/Down) [dB]: 5,5 / 11,0
SN Margin (Up/Down) [dB]: 6,0 / 26,0
Vendor ID (Local/Remote): TMMB / IKNS
Loss of Framing (Local/Remote):11 / 0
Loss of Signal (Local/Remote):1 / 0
Loss of Power (Local/Remote):0 / 0
Loss of Link (Remote):0
Error Seconds (Local/Remote):1 / 0
FEC Errors (Up/Down):0 / 0
CRC Errors (Up/Down):0 / 0
HEC Errors (Up/Down):0 / 0

It seems like its a big problem regarding BELL and TSI is aware of it, like Guspaz said I have tried disconnecting/reconnecting several times and it hasn't seemed to have worked, so I guess I am going to just wait it out unfortunately ....

quote from TSI STAFF

Hi,

We've identified that there are problems with DSL links as Bell is doing some changes on the network. It's going to sound odd, but right now the only way to get this fixed is to disconnect, reconnect, run a speed test and see if things have improved, if not, repeat the process. It's taking about 6 or so intentional disconnects force a move of the link your connected to. We're working to get the situation resolved in a permanent way, but right now this will get you to a less congested link as the maintenance finishes.

What is the sync speed showing in your GUI "connection" tab ?
On a Thomson it should be 192.168.1.254 to access the GUI menu.

Download DMT to determine your attenuation and distance to CO, it will also tell you what your line is capable of.

Download version DMT v7.35 for Thomson modems. Don't worry you can select English when the program installs ...LoL

»dmt.mhilfe.de/

I'm getting just under 3.9 Mbit/s downstream on a 6/0.8 Mbit/s sync. The downstream is normally about 5.1 Mbit/s (like everyone else on that profile).

Yayyyyyyyyy

From what I can see, seems like BHell have you on a 6Mbs/800Kbps profile and you are getting the 6dB upstream constantly which suggest it is the IKNS bug !

I kept reconnecting over and over again and did eventually get back to normal:



Still no upstream upgrades :P

I assume you refer to this speed.
Bandwidth (Up/Down) [kbps/kbps]: 512 / 6.016

Yes, sync speed is one indication, but SNR is another.
I noticed on your Thomson they have you are on interleave, OMG that is horrible ! ...LoL

Once this settles, you should ask TSI to switch you to Fast Path !

EDIT: here we go, you my friend definitely have the IKNS bug, BHell has you on a remote IKNS DSLAM and they probably misdiagnosed your line and put you on interleave !

Vendor ID (Local/Remote): TMMB / IKNS

back from running a DMT....
unfortunately it does not support firmware 8.2.7.7.0 , I can see DMT is a powerful tool, but too bad cannot use it.

thx EdT for your help, I guess I have to start a thread in direct section to hopefully get the problem solved.

One thing still confuse me, now I swap back to Bell 2wire modem again, and.....it's back to normal speed.


in 2Wire MDC I found this DSL Line (Wire Pair): Line 1 (inner pair)
Protocol: G.DMT Annex A
DSL Channel: Interleaved
DSLAM: Country: {0xB5} Vendor: {IKNS} Specific: {0x01}
ATM PVC Info: 0/35
ATM Encapsulation: LLC

It's also Interleaved....and I am still digging other info. in it.

Hrm, tonight my speeds were normal but no updates to my ticket. There was no slow down for me in Montreal. Shrug... I hope it stays like this!

now, to me it becomes a modem issue.....
try to avoid the randomness of one works and the other doesn't, I sway my two modems back and forth, but every time I got the same result.

2wire modem works at normal speed
TG784 works at 2M top......

If you read the TSI quote and TSI post above, it seems the low speeds are due to Bell messing with their network configuration and re-connecting the PPPoE session to (hopefully) get on a different link routing from your BAS to TSI's ERX is the only way to "fix" it. TSI's post says this requires about six attempts on average.

Maybe you got lucky and got on a good route when your PPPoE session came up on the 2wire, which made it seem like switching modems "fixed" it.

After doing the reset trick on my modem, for the time being all seems back to normal on my 6M service.

Yes, different types of DSL modems uses different chipsets and behave differently with the IKNS bug, but what is funny is that the 2Wire 2701 is known to perform badly with the IKNS and the Thomson is the one that is suppose to be better !

TSI only stated that they were having problems due to Bell shuffling things around on their end and that reconnecting to end up on a different AHSSPI seems to solve it, not that they actually had congestion issues at their end of things.

In any case, the problems are "initiated" from Bell's side of things so TSI has very little control over what happens.

I tend to agree with this theory that I just got lucky when I made the swap, coz the last time I tried before I went to bed last night I made a final swap to 2wire, and that time I only got 2.6M something.

but during the day time everything just works fine.

So to me the bottom line is we know the cause of the problem and we can expect an end and everything will be back to normal sometime

No. Quote from TSI verbatim :So...they acknowledge congested links and clearly suggest a workaround to help mitigate winding up on a congested link.