Stability options and enhanced traffic weighting

[#1]

It gets more and more peculiar:
I went from DGTeam 5.01.12 to both Netgear v5.01.01 and 5.01.14
With both Netgear firmware, it comes up with ADSL2+, but even lower speed and higher margin:

ADSL Link             Downstream     Upstream 
Connection Speed         1495 kbps 	         447 kbps 
Line Attenuation               43.5 db              23.2 db 
Noise Margin                    29.5 db              23.3 db 
               Down            Up
 G.dmt framing
K:              47(0)           16
R:              16              16
S:              1               4
D:              64              4
 ADSL2 framing
MSGc:           60              64
B:              46              15
M:              1               4
T:              1               1
R:              16              16
S:              0.9843          4.2953
L:              512             149
D:              64              4

If I force it to G.DMT, then I get the same as the DGTeam fw: 3072 kbps down and 448 kbps up

[#2]

even if i force g.dmt mode on mine it still syncs around 800kbps up your noise margin on the down is sky high and the 3072 sync looks like on of those banded syncs am on at the mo due to stabilty issues ive had now been upped to a 15db noise margin target

[#3]

Quote from: ASBO

.... looks like on of those banded syncs am on at the mo due to stabilty issues ive had ...

That's quite possible:
I left the router in multimode and it has gone from 1495 kb/s to 2272 kb/s down, and I've seen 445, 835 and now back to 448 kb/s up - this is with the Netgear firmware:
03/07: 1495/447
03/07: 1495/448
04/07: 1495/445
09/07: 2272/835 (probably changed on 06/07 as my profile had already gone to 2Mb/s by 09/07)
10/07: 2272/448
Now, with the change in sync to 2272 kb/s down, I checked again in G.DMT and it gave me the same speed - whereas with the 1495, I was getting double that in G.DMT.
With the change, I thought I'd try the DGT firmware, and it now works again, giving the same speed as with Netgear f/w - all very odd.
currently:

Mode:                   ADSL2+
Channel:                Fast
Trellis:                U:OFF /D:ON
Line Status:            No Defect
Training Status:        Showtime
                Down            Up
SNR (dB):       24.7            23.5
Attn(dB):       43.0            23.1
Pwr(dBm):       20.0            12.2
Rate (Kbps):    2272            448

Although it says 'fast', it's actually interleaved - I'm not sure why it does that as I also seen it reporting 'interleaved'  sometimes.

[#4]

Hiya,
Would you like me to send out a Thomson to see whether or not that improves performance at all?
It's certainly peculiar that the SNR is quite so high and I would have thought that if it were a banded profile, it would be the same on all pieces of kit, so would be good to see how it behaves on something different.

[#5]

Hi James,
Yes - I'll give the Thompson a try if you would like.
I did give my other router whirl last night:

Us Rate (Kbps)	 	448
Ds Rate (Kbps)	 	2268
US Margin	 	23
DS Margin	 	24
Trained Modulation	 	ADSL2Plus
DS Line Attenuation	 	38
US Line Attenuation	 	24
Path Mode	 	Interleaved

and forcing G.DMT mode:

Us Rate (Kbps)	 	448
Ds Rate (Kbps)	 	2272
US Margin	 	28
DS Margin	 	29
Trained Modulation	 	GDMT 
DS Line Attenuation	 	36
US Line Attenuation	 	25
Path Mode	 	Interleaved

and this is the DG834 this morning:

Mode:			ADSL2+
Channel:		Fast
Trellis:		U:OFF /D:ON
Line Status:		No Defect
Training Status:	Showtime
		Down		Up
SNR (dB):	24.9		23.2
Attn(dB):	43.0		23.1
Pwr(dBm):	20.0		12.0
Rate (Kbps):	2272		447

[#6]

Cheers- I've just ordered one which should be with you in the next couple of days.

[#7]

Thomson unit turned up.  Haven't had time to poke around too much, just set it up and left it running overnight.
Stats as of this morning at about 7 AM:

Uptime:	0 days, 9:07:49
DSL Type:	G.992.5 annex A
Bandwidth (Up/Down) [kbps/kbps]:	445 / 2,272
Data Transferred (Sent/Received) [kB/kB]:	0.00 / 0.00
Output Power (Up/Down) [dBm]:	12.0 / 19.5
Line Attenuation (Up/Down) [dB]:	23.5 / 41.5
SN Margin (Up/Down) [dB]:	20.5 / 26.5
Vendor ID (Local/Remote):	TMMB / TSTC
Loss of Framing (Local/Remote):	0 / 0
Loss of Signal (Local/Remote):	0 / 0
Loss of Power (Local/Remote):	0 / 0
Loss of Link (Remote):	0
Error Seconds (Local/Remote):	0 / 0
FEC Errors (Up/Down):	0 / 700
CRC Errors (Up/Down):	0 / 0
HEC Errors (Up/Down):	18,875 / 0

So, pretty much the same as the other two units.  Not sure why the 'Data transferred...' shows zero, is this a 'feature'?

[#8]

Not sure to be honest.
Would you like me to place an order to move you to "stable" to see what improvements you see, or do you want to play with the Thomson over the weekend?

[#9]

Hi James,
Probably leave it for the weekend - hopefully I can get a bit more acustomed to the Thomson and where to find stuff (e.g. remote access, mode selection, and so on...).
In terms of where we go from here, it depends to some extent whether you think it would be useful to give BT a poke with it as it is now - to get a better understanding of why it is doing what it's doing and what mechanisms you have to get resolution.  As a trial, that might give you some useful knowledge for if/when these options are offered more generally.  Two possible approaches come to mind:
a)  This user was generally at around 6 Mb/s on standard and has been dropped to 2 Mb/s on super-stable - is this to be expected, based on the line condition, or does it indicate a fault?
b)  This user has a minimum line speed of 5 Mb/s (or insert whatever the correct value is) and is currently at 2 Mb/s on super-stable - this should therefore be considered a fault for investigation.
Most of what we have seen is perfectly consistent with the system just being incredibly aggressive - leaving just the one period (about 28th June - 6th July) where I was getting the anomalous speeds between ADSL and ADSL2+ on one router, but that looks like something we'll never get to the bottom of (exchange end issue, user end issue, or some peculiar incompatibility between the two)...
On the other hand, if you think you got as much information as you need from my particular case, or you think the probability of getting anything useful out of BT is low - then I'm fine to make the transition to stable.
Let me know your thoughts, then we can agree the way forward next week.
Cheers
Geoff

[#10]

Geoff,
To be honest, I couldn't tell you what expected behaviour is, as you were our first trialist, apart from 1 member of staff on a very good line who didn't notice any changes.
I've asked Dave to feed back into BT to find out if what you have been experiencing is to be expected.

[#11]

Quote from: dusty_bin

(Thomson TG585v7 router) Not sure why the 'Data transferred...' shows zero, is this a 'feature'?

Yes, this is a 'feature'. We understand the information this data depends on isn't available in this router so the line should really be removed from the page. Perhaps someday Thomson will do that…
David

[#12]

Quote from: Jameseh

I've asked Dave to feed back into BT to find out if what you have been experiencing is to be expected.

Strange how this sort of things seems to happen...
...as I hadn't heard anything, I dropped Dave a PM early afternoon yesterday, and when I got home I found the speed had changed to:
Down: 5746 kbit/s - Margin: 15.5 dB
Up: 835 kbit/s
Quite an improvement from 2272/445 kb/s.
I didn't get a chance to extract all the link stats yet, and I'm guessing that this change probably happened yesterday morning, so wouldn't have been a result of my PM.
Cheers
Geoff

[#13]

I would love to be able to report some results after James switched me to the 'stable' option. However, before being switched (on Wednesday of last week) I was on a banded profile with my downstream sync speed limited to 3069 and an SNR of 14-16. After being unbanded and switched to 'stable' I had a day and a half where I was able to sync at around 5100 (with SNR=9). I then had my normal problem with SNR dropping rapidly to zero after dusk and the line becoming unusable after about 11pm.
Anyway, on Friday (17th) afternoon there was a resync forced from the BT end (for no obvious reason at that time, as SNR was holding steady at 9 in daytime) and since then I've been back on my old 3069/735/SNR14-16 combo, which I'm 99% sure means I'm back on the banded profile. So I don't think the DLM ever got to try any 'stable' settings - unless banding is part of the procedure.
The banded line sure is stable, though. With the higher frequencies essentially screened out, my SNR still drops a bit overnight, but not so much. However, a sync speed of 3069 (which means an IP profile of 2500 and a real throughput of around 2250) with an attenuation level of 39 is, to say the least, disappointing.
I'm attaching a graph of my post-dusk SNR drop on the day before I was 'rebanded' (actually it started a bit before dusk on that day, but it was dull and the street lighting, which is what I think is to blame, came on earlier). From previous experience I know the margin rises again after dawn. With a banded line the margin drops from 14-16 to 10-12 overnight.
Graham

[#14]

I should learn not to tempt fate...
Got home on Friday night to find I was back down to 3023 kb/s and mostly 23.5 SNR margin (sometimes down to 21.5 dB)
So I only got the higher rate for 1 day...
[edit] and in fact the uplink rate dropped too, probably on Saturday morning:

Uptime:	2 days, 12:33:11
DSL Type:	G.992.5 annex A
Bandwidth (Up/Down) [kbps/kbps]:	447 / 3,023
Output Power (Up/Down) [dBm]:	12.0 / 20.0
Line Attenuation (Up/Down) [dB]:	23.0 / 41.0
SN Margin (Up/Down) [dB]:	22.0 / 23.0

[#15]

Update showing that I was back up 6 Mb/s yesterday and now back to 3 Mb/s - which is similar to what we saw last week:
Transferred onto the 'SuperStable' stability option trial on 16/6/2009.
16/6 6615/1016 kbps SNRM  7.9 -  9.5 dB
17/6 6969/984 kbps  SNRM  7.4 - 11.3 dB
18/6 5539/1001 kbps SNRM 13.0 - 15.4 dB
19/6 2964/1001 kbps SNRM 20.2 - 22.5 dB
20/6 2964/855 kbps  SNRM 19.8 - 22.4 dB
22/6 2964/447 kbps  SNRM 21.3 - 22.6 dB
27/6 3511/447 kbps  SNRM 19.7 - 21.9 dB
28/6 1495*/447 kbps SNRM 27.6 - 29.6 dB
06/7 2272/835 kbps  SNRM 22.1 - 24.7 dB
09/7 2272/448 kbps  SNRM 22.3 - 24.8 dB
13/7 2272/447 kbps  SNRM 22.7 - 24.8 dB
17/7 2272/445 kbps  SNRM 25.0 - 27.0 dB
23/7 5746/835 kbps  SNRM 13.5 - 15.5 dB
24/7 3023/835 kbps  SNRM 21.0 - 23.5 dB
26/7 3023/447 kbps  SNRM 21.5 - 23.5 dB
28/7 2964/448 kbps  SNRM 19.5 - 21.5 dB
30/7 6655/447 kbps  SNRM  7.6 - 11.1 dB
31/7 3511/447 kbps  SNRM 22.3 dB
Notes:
1) Generally using Netgear DG834v4 except for the period 16/7 to 28/7 when Thomson 585v7 was used (figures shown for 28/7 are for DG834).
2) The 1495*/447 is the anomalous case reported earlier in the thread, when DGT firmware would not sync in ADSL2+ at all, standard f/w would sync as shown, and other routers would sync at around 3072 - as would DG834 in G.DMT mode with either f/w.
3) Interleaving applied throughout, except for the first day.  That first entry is also reasonably representative of the sync with 'Standard' stability before starting on 'Super Stable'.
4) Up to 27/6 DG834 SNR adjustment was set to 80%, later DG834 results are for 100%.

Quote from: Jameseh

I've asked Dave to feed back into BT to find out if what you have been experiencing is to be expected.

I did drop Dave a couple of PMs to see if he has had any feedback, and I have posted the results up to 28/7 in ticket ID: 29529941 for the attention of Dave and/or James.
The jumps just don't seem logical: regardless of the basis for making the changes, you would think DLM should try going up and down in 3 dB steps, rather than 12dB...
Cheers
[edit] 
Just looking at the logs timed at 12:03 and 16:03 on 30/07/2009, I can see that:
RSUnCorr went from 738 to 62728
ES: 39 to 327
SES: 1 to 11
etc.
So it could well be that the DLM is responding to something like that.  It still doesn't explain why it needs to make such large adjustments, though...