cancel
Showing results for 
Search instead for 
Did you mean: 

Connection still awful at times

Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

Quote from: myredhotcar
Something isn't right...

Hi Myredhotcar,
You need to be aware that sometimes test equipment (programs) can misreport what is happening.  Unless there is corresponding evidence of synch loss (for example router logs or a RADIUS connection graph) the graph shown on that post suggest to me the routerstats program was not able to retrieve numerous router data samples (form time to time) and thus plotted zero values.  I have extensive experience in running routersats and see this a lot in conjunction with the TG routers.  When this happens, the SUMMARY tab might show anything but the DSL stats.  It could be the tone buckets, known routing table to name but two.
I have had email exchanges with John the author of RS and he advises me that one should never see anything other than DSL info on the summary page.  If one does, it is indicative of RS having lost some internal windows messages.  Clicking the stop and start capture buttons "fixes" the issue.
What do you think is wrong with your BB experience?    Is it the RS graphs alone which you are using as an indication that the link has disconnected frequently?
On a 40dB line your synch rate is fairly high (http://www.kitz.co.uk/adsl/max_speed_calc.php suggests 9Mbps at a SNRM of 6dB) and thus at 10.6Mbps error counts might well be expected.  I have experienced TG routers simply not responding at all to the admin interfaces when encountering a high rate of FEC management with the consequence that my graphs looked like yours, though at the time a possibly defective router / PSU was suspected and removed from the equation by taking a replacement.
Quote from: Chris
It doesn't mean that your SNR won't drop below 9db, this just means that your SNR Target won't go down to 6db or 3db. DLM would normally, on a healthy line, do this to give you the best speeds that your line can handle whilst maintaining stability.
SNR  Explained:
... The idea of SNR is that it will fluctuate depending on your line conditions. It's the SNR which will increase decrease if there is interference on your line, for example. we can't control the interference on your line nor the SNR that your line is set working at. we can set the target but we can't force it to stay static at 9db for example. If we could do that, your line would constantly drop out if it needed to fluctuate upwards downwards.
If your SNR is fluctuating then we need to find out what the source of this is. SNR doesn't fluctuate on it's own, something is more than likely causing interference on your line - what this is, I couldn't say as it could be something within your premises or something causing interference between the outside of your property all the way to the Local Exchange.

I felt a need to clarify Chris' statement or I've totally misread it...
The TARGET SNRM (which is what you requested to be set to 9dB) is used at synch-up time between the modem and the exchange which seek to achieve the highest possible speed at that SNRM given the then current line noise.
That synch speed stays static for the duration of the DSL session.  If the background noise decreases, the SNRM goes up, if the noise increases it goes down.  If it goes down too low, synch will be lost and the connection dropped.  In such circumstances you are likely to get a new (lower) synch speed at the TARGET SNRM.
If you are experiencing an unstable line, there is merit in forcing a resynch after dusk - this will be at a reduced synch rate, but hopefully one which will perform comparatively better due to there being fewer errors requiring correction.

HTH.

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

Pettitto
Plusnet Alumni (retired)
Plusnet Alumni (retired)
Posts: 6,346
Fixes: 5
Registered: ‎26-11-2011

Re: Connection still awful at times

I'm not sure that the amendments you've made the quote are correct or we're looking at SNR/DLM from a different stand point.
When a connection drops out, due to interference for example, the SNR will increase, from say, 3db to 9db - DLM may increase the SNR Target from 3db to 6/9db to help keep the line stable.
myredhotcar
Grafter
Posts: 457
Registered: ‎06-11-2013

Re: Connection still awful at times

@ Chris Pettitt: many thanks  Smiley
@ Townman: thanks for the reply, how do I access my RADIUS connection graph? I seem to remember accessing it back when I first joined but cannot for the life of me find it now. I would be interested to see them anyhow.
I joined PlusNet in November and I am currently on my 3rd router; I started with a DG834G v3 and this is what the graphs looked like with that router back then:

The line was, at this point, utterly unusable (constant loss of synch). DCT advised me that the DG834G v3 may have been the culprit, so I switched to a cheap spare router (WBR-T2) while awaiting a Broadcom based Netgear. I never managed to get that router logging with RouterStats but here are some grabs from the logs of that router:
Nov 11 21:18:30    syslog    emerg    BCM96345 started: BusyBox v1.00 (2007.10.20-02:18+0000)
Nov 11 21:18:30    user    notice    kernel: klogd started: BusyBox v1.00 (2007.10.20-02:18+0000)
Nov 11 21:18:41    user    warn    kernel: blaadd: open error -125
Nov 11 21:18:53    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:18:53    user    warn    kernel: blaadd: open error -125
Nov 11 21:19:05    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:19:05    user    warn    kernel: blaadd: open error -125
Nov 11 21:19:17    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:19:17    user    warn    kernel: blaadd: open error -125
Nov 11 21:19:29    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:19:29    user    warn    kernel: blaadd: open error -125
Nov 11 21:19:41    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:19:41    user    warn    kernel: blaadd: open error -125
Nov 11 21:19:53    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:19:53    user    warn    kernel: blaadd: open error -125
Nov 11 21:20:05    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:20:05    user    warn    kernel: blaadd: open error -125
Nov 11 21:20:17    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:20:17    user    warn    kernel: blaadd: open error -125
Nov 11 21:20:29    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:20:30    user    warn    kernel: blaadd: open error -125
Nov 11 21:20:42    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:20:42    user    warn    kernel: blaadd: open error -125
Nov 11 21:20:54    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:20:54    user    warn    kernel: blaadd: open error -125
Nov 11 21:21:06    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:21:06    user    warn    kernel: blaadd: open error -125
Nov 11 21:21:18    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:21:18    user    warn    kernel: blaadd: open error -125
Nov 11 21:21:30    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:21:30    user    warn    kernel: blaadd: open error -125
Nov 11 21:21:42    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:21:42    user    warn    kernel: blaadd: open error -125
Nov 11 21:21:54    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:21:54    user    warn    kernel: blaadd: open error -125
Nov 11 21:22:06    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:22:06    user    warn    kernel: blaadd: open error -125
Nov 11 21:22:18    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:22:18    user    warn    kernel: blaadd: open error -125
Nov 11 21:22:30    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:22:30    user    warn    kernel: blaadd: open error -125
Nov 11 21:22:42    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:22:42    user    warn    kernel: blaadd: open error -125
Nov 11 21:22:54    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:22:54    user    warn    kernel: blaadd: open error -125
Nov 11 21:23:06    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:23:06    user    warn    kernel: blaadd: open error -125
Nov 11 21:23:18    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:23:18    user    warn    kernel: blaadd: open error -125
Nov 11 21:23:30    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:23:30    user    warn    kernel: blaadd: open error -125
Nov 11 21:23:42    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:23:42    user    warn    kernel: blaadd: open error -125
Nov 11 21:23:54    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:23:54    user    warn    kernel: blaadd: open error -125
Nov 11 21:24:06    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:24:06    user    warn    kernel: blaadd: open error -125
Nov 11 21:24:18    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:24:18    user    warn    kernel: blaadd: open error -125
Nov 11 21:24:31    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:24:31    user    warn    kernel: blaadd: open error -125
Nov 11 21:24:43    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:24:43    user    warn    kernel: blaadd: open error -125
Nov 11 21:24:55    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:24:55    user    warn    kernel: blaadd: open error -125
Nov 11 21:25:07    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:25:07    user    warn    kernel: blaadd: open error -125
Nov 11 21:25:19    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:25:19    user    warn    kernel: blaadd: open error -125
Nov 11 21:25:31    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:25:31    user    warn    kernel: blaadd: open error -125
Nov 11 21:25:43    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:25:43    user    warn    kernel: blaadd: open error -125
Nov 11 21:25:55    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:25:55    user    warn    kernel: blaadd: open error -125
Nov 11 21:26:07    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:26:07    user    warn    kernel: blaadd: open error -125
Nov 11 21:26:19    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:26:19    user    warn    kernel: blaadd: open error -125
Nov 11 21:26:31    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:26:31    user    warn    kernel: blaadd: open error -125
Nov 11 21:26:43    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:26:43    user    warn    kernel: blaadd: open error -125
Nov 11 21:26:55    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:26:55    user    warn    kernel: blaadd: open error -125
Nov 11 21:27:07    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:27:07    user    warn    kernel: blaadd: open error -125
Nov 11 21:27:19    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:27:19    user    warn    kernel: blaadd: open error -125
Nov 11 21:27:31    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:27:31    user    warn    kernel: blaadd: open error -125
Nov 11 21:27:43    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:27:43    user    warn    kernel: blaadd: open error -125
Nov 11 21:27:55    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:27:55    user    warn    kernel: blaadd: open error -125
Nov 11 21:28:07    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:28:07    user    warn    kernel: blaadd: open error -125
Nov 11 21:28:19    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:28:19    user    warn    kernel: blaadd: open error -125
Nov 11 21:28:32    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:28:32    user    warn    kernel: blaadd: open error -125
Nov 11 21:28:44    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:28:44    user    warn    kernel: blaadd: open error -125
Nov 11 21:28:56    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:28:56    user    warn    kernel: blaadd: open error -125
Nov 11 21:29:08    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:29:08    user    warn    kernel: blaadd: open error -125
Nov 11 21:29:20    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:29:20    user    warn    kernel: blaadd: open error -125
Nov 11 21:29:32    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:29:32    user    warn    kernel: blaadd: open error -125
Nov 11 21:29:44    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:29:44    user    warn    kernel: blaadd: open error -125
Nov 11 21:29:56    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:29:56    user    warn    kernel: blaadd: open error -125
Nov 11 21:30:08    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:30:08    user    warn    kernel: blaadd: open error -125
Nov 11 21:30:20    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:30:20    user    warn    kernel: blaadd: open error -125
Nov 11 21:30:32    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:30:32    user    warn    kernel: blaadd: open error -125
Nov 11 21:30:44    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:30:44    user    warn    kernel: blaadd: open error -125
Nov 11 21:30:56    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:30:56    user    warn    kernel: blaadd: open error -125
Nov 11 21:31:08    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:31:08    user    warn    kernel: blaadd: open error -125
Nov 11 21:31:20    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:31:20    user    warn    kernel: blaadd: open error -125
Nov 11 21:31:32    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:31:32    user    warn    kernel: blaadd: open error -125
Nov 11 21:31:44    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:31:44    user    warn    kernel: blaadd: open error -125
Nov 11 21:31:56    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:31:56    user    warn    kernel: blaadd: open error -125
Nov 11 21:32:08    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:32:08    user    warn    kernel: blaadd: open error -125
Nov 11 21:32:20    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:32:21    user    warn    kernel: blaadd: open error -125
Nov 11 21:32:33    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:32:33    user    warn    kernel: blaadd: open error -125
Nov 11 21:32:45    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:32:45    user    warn    kernel: blaadd: open error -125
Nov 11 21:32:57    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:32:57    user    warn    kernel: blaadd: open error -125
Nov 11 21:33:09    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:33:09    user    warn    kernel: blaadd: open error -125
Nov 11 21:33:21    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:33:21    user    warn    kernel: blaadd: open error -125
Nov 11 21:33:33    daemon    notice    pppd[555]: PPP: Start to connect ...
Nov 11 21:33:33    user    warn    kernel: blaadd: open error -125

At this point the line was barely usable (both routers kept constantly losing synch although the Broadcom based WBR-T2 did a little better than the DG834G v3).
With regard to what I think is wrong with my BB experience, when the noise margin drops below around 6dB:
pages load VERY slowly (as if I am using dial-up)
pages time out
pages will load and images may not (I have checked MTU settings, currently at 1432)
if I am running Transmission torrents drop to 0 kB/s until the noise margin picks up
I cannot use video streaming services
I cannot log into my bank or email
and finally, during these periods, if I run a BTW speed test I get results such as this:

I don't continually monitor my line, I only run RouterStats or DSLstats and speed tests when there are issues, or if I have been told to monitor the line.
When you say there may be merit forcing a resynch at dusk do you mean force a reconnect in the router web interface?
As mentioned, this is the 3rd router I have tried on my line, so it seems suspicious that my equipment is at fault. Statistically it is a possibility that I have had 3 bad routers in a row (although both the GT and the WBR-T2 were brand new) but I guess the odds for this are fairly low.
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

Quote from: Chris
I'm not sure that the amendments you've made the quote are correct or we're looking at SNR/DLM from a different stand point.
When a connection drops out, due to interference for example, the SNR will increase, from say, 3db to 9db - DLM may increase the SNR Target from 3db to 6/9db to help keep the line stable.

Chris,
Sorry for that, I think it depends on how you read the question / answer...
I read the question as "Why if my SNRM has been set to 9dB, does it change to some other value [as viewed by router stats] and not stay at 9dB as has been set by PN?".
There have been a number of threads over time conveying an expectation that once the (target) SNRM has been set to (say) 9dB that this the only value people expect to see in their router stats, which is what I thought you were explaining why this is not the case.  You were not, you was explaining how if noise persists so as to cause disconnections the TARGET SNRM will be increased by the DLM at the next reconnection in an attempt to maintain a connection in the presence of the noise.
It is the important difference between TARGET SNRM (as set by the DLM) and CURRENT SNRM being seen by the router.
I think that both versions stand, depending upon which question is being asked.
Cheers,
Kevin

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

myredhotcar
Grafter
Posts: 457
Registered: ‎06-11-2013

Re: Connection still awful at times

I was wondering why, when my line was set to 9dB, it dropped below 9dB. If that helps.
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

Quote from: myredhotcar
how do I access my RADIUS connection graph?

One of the helpful DCT guys publish it for you here.  That does take a bit of their time, what I wanted to be able to do from that is give your a means of confirming a time correlation with the SNRM graph dips with actual disconnect logs to determine if they were real or a phantom of the measuring system (RouterStats).
I cannot read the graphs you've inserted, attaching them can be more practical.
Quote from: myredhotcar
...so I switched to a cheap spare router (WBR-T2) while awaiting a Broadcom based Netgear. I never managed to get that router logging with RouterStats but here are some grabs from the logs of that router:
At this point the line was barely usable (both routers kept constantly losing synch although the Broadcom based WBR-T2 did a little better than the DG834G v3).

Whatever was going on there, it is clear that PPP was not getting up, let alone being dropped.
Quote from: myredhotcar
With regard to what I think is wrong with my BB experience, when the noise margin drops below around 6dB:
pages load VERY slowly (as if I am using dial-up)
pages time out
pages will load and images may not (I have checked MTU settings, currently at 1432)
if I am running Transmission torrents drop to 0 kB/s until the noise margin picks up
I cannot use video streaming services
I cannot log into my bank or email
and finally, during these periods, if I run a BTW speed test I get results such as this:
I don't continually monitor my line, I only run RouterStats or DSLstats and speed tests when there are issues, or if I have been told to monitor the line.
When you say there may be merit forcing a resynch at dusk do you mean force a reconnect in the router web interface?
As mentioned, this is the 3rd router I have tried on my line, so it seems suspicious that my equipment is at fault. Statistically it is a possibility that I have had 3 bad routers in a row (although both the GT and the WBR-T2 were brand new) but I guess the odds for this are fairly low.  

Cannot read the BTw report.

This is ugly!  When the SNRM as reported by the router drops, it indicates that there is external interference giving rise to errors and the need to perform error recovery.  This is a useful explanation if you are interested in technical details - http://www.kitz.co.uk/adsl/lowSNR.htm
Have you looked at your internal phone cabling?  Do you have any time correlation to the issues - is it worse at one time of the day than another?
At this point I would recommend  running RS 24x7 to see if there are any patterns to this and if possible avoid restarting the router... when I say force a resynch, I mean a clean start of the router.  You ought though (as you rather suggest) force a disconnect through the router interface.  This will avoid the DLM messing with the TARGET SNRM as per Chris' original explanation.
Three duff routers, probably not, but I went through 4 before we (Chris and I) profiled the problem to be a number of external REIN issues.

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

Pettitto
Plusnet Alumni (retired)
Plusnet Alumni (retired)
Posts: 6,346
Fixes: 5
Registered: ‎26-11-2011

Re: Connection still awful at times

Here's the graph
<img src="http://community.plus.net/visualradius/generated/image13993907684311.png"/>
I think we've both got the same idea but we've looked at it a different way  Crazy
myredhotcar
Grafter
Posts: 457
Registered: ‎06-11-2013

Re: Connection still awful at times

Hi townman, sorry that you couldn't read the graphs, they can all be found on the following thread http://community.plus.net/forum/index.php/topic,119868.0.html which was the first thread I started.
My house is around 9 years old so there should be no archaic wiring; I have nothing but my router plugged in (no extensions etc) and this has been in the test socket since I first started getting issues in November. The issues seem to slowly get worse over the course of the day although there have been periods where things have been fine.
To my (totally untrained) eye it seems that each day the noise margin had dropped a little; as a result of the previous fault the SNR was reset and gradually dropped to just below 3dB. It usually seems okay if it stays above 6-7dB but DLM pushes it lower and lower. Things will be fine for a period and I suddenly start getting issues. Things have been particularly bad of late although okay so far today. The only possible correlation I have noticed is wet weather, but in the thread I linked you to I included RouterStats grabs from close neighbours and they don't have the same issues so it seems the issue lies on my line itself. Again, I am untrained so essentially guessing but I am suspicious that there is a problem with the line to my house, possibly where it comes in. I say that but every quiet line test I have done is totally quiet, so plucking at straws a bit.
Things I hope I have ruled out:
Alarm systems (there are none nearby)
heating system (heating has been off for several weeks now)
microwaves (I use the microwave maybe twice a week)
EPOS and PDQ devices (around 150 metres from the closest shop)
I have no television or set top box and very few electrical items
How did you troubleshoot REIN? Would it be prudent to unplug everything I own and see if there is any difference, before plugging things in one at a time or is that OTT or pointless?

@Chris Pettitt: thanks for the graph, that looks a lot better than I expected. One thing I have noticed is that the router is making a regular clicking sound at times, almost like the sound of a relay. The WBR-T2 also did this, although I thought I was imagining it. It seems slightly louder with the Netgear. Sometimes it is very regular, at other times not so. The power supplies are different BTW so I can hopefully rule that out. The only things unchanged are the line itself, and also all three routers have been plugged into the same mains socket. There is a two way mains socket and I have tried both sockets; could it be the socket itself?
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

@Chris,
It is nice to agree!
Thank you for the RADIUS graph, it confirms what I was suspected, well at least for Sunday evening.

@myredhotcar,
The SNRM graph you posted from post 23:00 on Sunday night which seemed to suggest a lot of line drops, is most likely to be a RouterStats reporting issue.
I say "most likely" because the RADIUS graph does not show any PPP session losses after 22:00 on Sunday.  I know that it is possible to loose synch yet maintain PPP - at least I am convince that I have see it - but it is rare.
What cannot be confirmed or speculated on is the reason for the reporting issues - either the router did not respond to stats requests or if RS had got its message sequence screwed up.

Can you please do the following...
1. Get router stats running 24x7 - 10 second sample interval, 720 points per graph - this gives a 2 hour "window"
2. Plot RX speed on the RX SNRM graph - this halves the number of graphs you need to publish
3. On all graphs, click "make available to webserver" - that creates historical views as well as a "current" view in the RS folder.
4. Plot one of the error counts on the custom graph, set to "count since last sample"
A 10 second sample time will allow you to see more details in SNRM shifts and the detailed shape of CRC counts, which if this is a REIN issue can help point to the source.
If this is REIN, be warned, it can be very hard to find and even then impossible to eliminate - BTOR have no obligations in respect of REIN.  Those which do exist lay with Ofcom.
Will watch this one with interest.

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

myredhotcar
Grafter
Posts: 457
Registered: ‎06-11-2013

Re: Connection still awful at times

Hi townman thanks for your assistance, I really appreciate it.
I (think) I have RouterStats set up as you describe, except 4. Plot one of the error counts on the custom graph, set to "count since last sample". I cannot for the life of me work out how to plot the error count.
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

Quote from: myredhotcar
sorry that you couldn't read the graphs, they can all be found on the following thread http://community.plus.net/forum/index.php/topic,119868.0.html which was the first thread I started.

The graph here - http://community.plus.net/forum/index.php/topic,119868.msg1039551.html#msg1039551 - is most interesting.  Do you by any chance live near an electric train line - specifically the WCML?  That graph shows a reasonably steady (good SNRM) and spiked SNRM dips.  Compare that to the attached...  The spikes are WCML super-pendileno trains passing my (then) d-side 500m metres away.  If you think that this might be relevant, look here http://realtraintimes.org.uk - go to the detail pages and find a location near to you.  You might need to look at the next station up or down the line first.  For example, I was near Weaver Junction, which is just North of Acton Bridge Station.  From there look at the detailed timing of a specific train to find a timing point nearer to you.  Look for passing trains and see if they correlate with your spikes.  Not all trains will give the issues (if this is what it is) it is only class 390 EMUs.
Quote from: myredhotcar
To my (totally untrained) eye it seems that each day the noise margin had dropped a little; as a result of the previous fault the SNR was reset and gradually dropped to just below 3dB. It usually seems okay if it stays above 6-7dB but DLM pushes it lower and lower.

I need to be sure that I understand exactly what you are saying here.  Do you mean that you see the SNRM simply getting lower over time WITHOUT a resynch or are you saying that through a number of resynchs the SNRM is STARTING after the resynch lower than it was?
I suspect you mean that the SNRM you see when monitoring is getting lower and lower WITHOUT a resynch taking place.  If that is correct, then that is what I was trying to explain when I changed Chris' words.  In the presence of noise the reported CURRENT SNRM will go down - indeed if it gets below 3dB there is a marked risk that the link will drop, if not it will certainly slow down whilst all of the error management (CRC corrections and retransmissions) take place.  If you can confidently identify this scenario, it would be an excellent time to do the forced restart I suggested, as this will give you a (slower) synch rate at the target SNRM.  When the noise vanishes, you might well see the CURRENT SNRM to be reported as 12dB or even 15dB.

Quote from: myredhotcar
The only possible correlation I have noticed is wet weather

Wet weather can point to a joint problem and such lines can be more susceptible to REIN - @DCT what does the copper pair balance look like please?  Joint problems can be difficult to find if it is dry when the BTOR engineer is looking!
Quote from: myredhotcar
Things I hope I have ruled out:
Alarm systems (there are none nearby)
heating system (heating has been off for several weeks now)
microwaves (I use the microwave maybe twice a week)
EPOS and PDQ devices (around 150 metres from the closest shop)
I have no television or set top box and very few electrical items
How did you troubleshoot REIN? Would it be prudent to unplug everything I own and see if there is any difference, before plugging things in one at a time or is that OTT or pointless?

Those typical sources might not actually be in your home!
No that is not OTT / pointless - there is every possibility that the cause is not obvious.  Having moved back home from the business premises (to which the graph related) I had a bit of tidying up to do here after our daughter left.  I removed a not-used PSU plugged into the mains socket next to the active router and attained a 1.5dB improvement on the SNRM.  Clearly on the next resynch, that delivered an improved synch speed.
Locating REIN is difficult.

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

Oldjim
Resting Legend
Posts: 38,460
Thanks: 787
Fixes: 63
Registered: ‎15-06-2007

Re: Connection still awful at times

Open Routerstats Help
Click on Using the graphs then Technicolor TG582/587 and it tells you how
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

OldJim,
Myredhotcar is not using the TG routers.
I know that Jon who wrote RouterStats (please confirm that this - http://www.vwlowen.co.uk/internet/files.htm - is what you are using) put a lot of extra bells and whistles in for NG routers.  Looking at these graphs again, ther are a little different to what I see against TG routers.  In that configuration, one has to select some metric from the summary page (a counter) for be plotted on the custom graph.
These screen shots have an errors tab within which there is a CRC and a FEC graph.  As I have no direct experience of this configuration, my guidance might be flawed.
On the TG configuration, there is a right click on graphs option list, one of which is a "plot difference between each sample" option.
BTW myredhotcar, I have just seen my RS plotting zeros whilst displaying (in error) the bits per tone data on the summary tab.

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.

myredhotcar
Grafter
Posts: 457
Registered: ‎06-11-2013

Re: Connection still awful at times

When I click help I only get a single page of information, I have visited the online help and am still confused. This thread and this page seem to suggest the answer may lie in the 'Router Search Text' tab but I cannot work it out as of yet. I have tried finding the relevant section using 'source A' and 'source B' to no avail. The fact that the terminology is alien doesn't help. I will keep reading and will hopefully find the answer. I am probably missing something simple.
With regard to electric train lines my closest station is Bradford on Avon (Wessex Main Line) which isn't electrified. According to Google Maps I am 0.344 km away from the nearest point of the line, as the crow flies.
To clarify, the SNRM seems to get lower without a SNR reset; it seems that is what it does. It slowly got lower and lower after I joined, once things were stable, and a member of the DCT agreed to set the target to 9dB and performed a SNR reset. Things were stable but the SNRM slowly dropped to below 3, hence this thread. Unfortunately I am not very familiar with the terminology so I may be making mistakes.
I forced a disconnect (21:37) and then reconnected at around 21:39, using the connection status popup in the web interface of my router. Pages wouldn't load but RouterStats carried on monitoring the line with no apparent change in any of the graphs. Is this normal behaviour?
Tomorrow I will unplug everything from all sockets and see if there is any difference. I will then plug everything back in one item at a time to see if there is any bearings on the graphs. I think I could do with reading about REIN a bit more also.
Things have been okay today but pages have been taking a long time to load in the last 90 minutes or so. It is usable though.
Cheers Townman
Townman
Superuser
Superuser
Posts: 22,919
Thanks: 9,536
Fixes: 158
Registered: ‎22-08-2007

Re: Connection still awful at times

Quote from: myredhotcar
When I click help I only get a single page of information, I have visited the online help and am still confused. This thread and this page seem to suggest the answer may lie in the 'Router Search Text' tab but I cannot work it out as of yet. I have tried finding the relevant section using 'source A' and 'source B' to no avail. The fact that the terminology is alien doesn't help. I will keep reading and will hopefully find the answer. I am probably missing something simple.
Things have been okay today but pages have been taking a long time to load in the last 90 minutes or so. It is usable though.

IIRC you are using a NG router - RouterStats works rather differently for those routers.  The A & B sources you are referring to relate to other routers.  I find the easiest way to configure the custom graphs is to go to the SUMMARY tab.  In the US or DS window, select the 'label' text next to a counter, right click the mouse on the [user-def X] option.  It would be good to see actual or confirmation that there are no error events displayed on the error graphs.
The latest graphs look quite clean, as for slow loading, is that persistently or intermittently?  This is quite a change around from those plots you've uploaded before... dammed mystery!
I think it might be useful to (whilst the line is in this state) to recollect all of the performance data - Router Data, BTw speed test (including profile) and PN profile (Current Speed).

Kevin

Superusers are not staff, but they do have a direct line of communication into the business in order to raise issues, concerns and feedback from the community.