Bicephale
join:2005-09-24
 ·TekSavvy Solutions..
1 edit | reply to morphector
Re: In-depth analysis of line stats
Yes! That's the one all right!!!
I wish i could learn a few good facts about this resource!...
For example, what's the maximum duration of such records.
I'm even more curious now. Would you know if parameters
known as Impulse Noise Protection (INP) and/or Low Depth
Interleave can be found somewhere in the "Lantern" pages?
 |
|
STN
join:2004-12-29
Canada
·Bell Sympatico
1 edit | said by Bicephale  :For example, what's the maximum duration of such records. Depends on DSLAM. Max is around
24hrs in 15min intervals
7 days in 24hr intervals
said by Bicephale :I'm even more curious now. Would you know if parameters known as Impulse Noise Protection (INP) and/or Low Depth Interleave can be found somewhere in the "Lantern" pages? I don't believe so, but each profile has specific values of these parameters. So by knowing the profile (5056/800, etc) you could cross-reference those values. But without the profile info you can't see it from these pages. |
|
Bicephale
join:2005-09-24
·TekSavvy Solutions..
| Oh i see, fifteen minutes that's if one reads everything all
at once - but what if someone were to access that page
every minute? Would that individual read the same values
during one fifteen minutes interval or would he be able to
effectively reach a resolution of 1 minute?...
|
|
STN
join:2004-12-29
Canada
·Bell Sympatico
3 edits | I don't follow the question exactly, but I think you may misunderstand.
First off these performance counters do not report the current errors. What happens is the DSLAM starts a 15 minute timer. During this time he keeps track of all the errors and other counts that happen. This is "Current 15 Minutes". Once the 15 minutes is up, he rolls this over into a 15 minute container. That would be your first previous 15 min numbers, and it would reflect everything that happened in those 15 minutes.
Then once it rolls over it restarts the 15 minute timer and tracks all numbers starting from 0. Once that 15 minutes ends, it pushes the previous interval values (1) over one container. So container 1 numbers go into container 2, and the numbers that just came out of the "curr15min" now go into container 1. And curr15min is restarted again.
So at this point we have:
curr15min (which is counting from 0's now)
15min interval 1 (which is counts from the previous 15min)
15min interval 2 (which is counts from 15min that happened 30min ago)
This continues until 96 intervals are done, which is 24 hours. So you now have "24 hours of 15 minute interval data". When the next curr15min ends, it flushes out interval 96 data, moves everything over (which empties interval 1) and it puts the curr15min that just ended into interval 1.
So if I want to see what happened in the 15min that occurred 12 hours ago, I would review interval 48.
Also, when the original curr15min clock began another clock began: curr1day time. This counts up to 24 hours, and then rolls over. When errors occur and that is populated in the curr15min data it is also populated in curr1day data. Once 1day is up it puts the data into prev1day interval 1. Rince and repeat for 7 intervals.
So you also have "7 days worth of 24 hour data", or on some DSLAMs only 2 days worth of 24 hour data.
Also, the "curr15min" counts are available in other windows, but the windows above show the previous interval data. |
|
Bicephale
join:2005-09-24
·TekSavvy Solutions..
| I'm asking because DSLRocker could use this 2Wire (CPE) page:
New teksavvy customer slow download speed???, Bicephale, 2009-Nov-4
...to generate this higher resolution Rx CRC graphic:
%20.GIF)
Help/Advice with my Line Statistics please, Bicephale, 2009-Jul-8
The 3rd column is clearly titled "Current 15-Minute Interval" so
i thought if it works on CPE equipment then who knows what's
going to happen on a DSLAM (or GlobeSpan/IKNS remote) seen
via Bell's level-3 "Lantern"!...
 |
|
morphector
join:2009-04-06
Jonquiere, QC
| reply to Mauricio
I don't think you can get graphics of this kind via lantern 3 (correct me someone if i'm wrong)
as far as I know lantern level 3 over level 2 gives you:
- reseting of port and card (nice they don't give it to everyone because everybody would be reseting cards everytime so all customers would lose sync at any time)
-can change speed (they removed it to tech because 5-6-7 years ago only bell tech were doing high speed repairs, some of them never went to customers and were just lowering speed)
-can ''set up'' a port so it's identified with a phone number. |
|
Mauricio
aka CigaR
Premium
join:2008-12-06
1 edit | reply to Mauricio
It seems I was able to stabilize my line. Added a RF choke (6 turns) + I shortened the wire from the ADSL/POTS splitter to the modem.
I also ran a short cat5e from the demarc to the splitter. So far, it seems to be holding off quite well with no fluctuations in SNR. It has been steady @ 14dB. Modem is reporting about 120 CRC errors over a period of 25 minutes; not bad given my line condition.
It has to be monitored over a longer period though. The line seems quite "clean" today, with no noise impulses.
*crosses fingers* 
Funny thing is, if my line remains noise-free, my modem is reporting that it could achieve 2700 kbps @ 65 dB... Yeah right ! lol |
|
Bicephale
join:2005-09-24
·TekSavvy Solutions..
| Hi Mauricio,
I understand you have a loaded schedule so i appreciate your
feedback!... I don't remember what your MoDem brand/model
was exactly but if you could manage to leave either 'DMT' (in
"Diagnosis" mode), TerTech's 'DSL_StatScope' or 'RouterStats'/
'RouterStats-Lite' running for at least a couple days to collect
long-term Error Rate statistics, with and without an RF Choke
so that "Before"/"After" comparisons can be made, then i'd be
glad to handle the number crunching for you! Just plug/unplug
the RF Choke once while the tool(s) suitable for your MoDem
has been launched and meet me back here after a week or so
(or even more!) has passed if that's OKay for you... Averaging
should reveal differences in noise level if present by rejecting
marginal events (that's why it would be desirable that the test
lasts for as many days as possible): only the overall behaviour
gets "amplified" when daily records are averaged so the more
daily records you get the better. Overall this whole operation
isn't supposed to require more than half an hour of your time.
 |
|
dslrocker3
join:2002-05-26
Toronto
2 edits | reply to Mauricio
said by Mauricio :Funny thing is, if my line remains noise-free, my modem is reporting that it could achieve 2700 kbps @ 65 dB... Yeah right ! lol Oh, it'll connect at 2700Kbps. The 2wire attainable rate are usually pretty accurate. On the Broadcom-based unit, there's a bug in the attiainable rates where they are reported too high if you are on an interleaved profile (have never seen an RCO above 92% on an interleaved profile, and I'm not talking about misreported IKNS interleaved). I noticed no such discrpency in the 2wire modem's attainable rate calculations.
I'm very surprised that you appear to be still on fast path. On a line like yours, they don't always give you that option. If you have decided that interleaved would be fine for you, 2496/640 might work okay.
do you have access to any other modems (spares/borrow from neighbors or friends)? Just as I often say that the Speedtouch 516 is not the best on all lines, the same can be said about the 2wire 2700. When line conditions are bad, attenuated, or on a bridge taps, all types of strange things can happen such as large differences in modem performance among different models. |
|
%20.GIF "179949 bytes")
