Strange sync-time SNRM

For some time my line has flipped occasionally between 3dB and 4dB, with a recent 5dB, depending on local crosstalk and suchlike, then over time my attainable on both down and upstream has crept up.

Around the 23rd day after being set at 4dB DLM will resync the line with a 3dB margin.

However in the last six days when I had previously sync'ed on 4dB but the line had within minutes dropped to varying at 3.5dB and just below, (normal variation behaviour of course) I appear to have had two very rapid re-syncs, in that the connection is lost for a much shorter time than usual.

Each time the margin (shown on DSLStats graphs) has clearly again been at 3.5dB, but with significant sync increases.

The last three syncs, all with the margin sticking at 3.5dB and wobbling as low as 3.1dB, have been 65717/12459; 67212/12130 and 693199/12403. Attainable is always above actual and at the moment I have:-

xdslctl info --show
xdslctl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 12444 Kbps, Downstream rate = 71624 Kbps
Bearer: 0, Upstream rate = 12403 Kbps, Downstream rate = 69319 Kbps
Bearer: 1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 3.4 6.0
Attn(dB): 19.4 0.0
Pwr(dBm): 13.7 7.6

VDSL2 framing
Bearer 0
MSGc: -6 26
B: 243 237
M: 1 1
T: 0 51
R: 10 16
S: 0.1121 0.6100
L: 18129 3331
D: 8 1
I: 254 127
N: 254 254
Q: 8 0
V: 0 0
RxQueue: 76 0
TxQueue: 19 0
G.INP Framing: 18 0
G.INP lookback: 19 0
RRC bits: 0 24
Bearer 1
MSGc: 154 -6
B: 0 0
M: 2 0
T: 2 0
R: 16 0
S: 6.4000 0.0000
L: 40 0
D: 3 0
I: 32 0
N: 32 0
Q: 0 0
V: 0 0
RxQueue: 0 0
TxQueue: 0 0
G.INP Framing: 0 0
G.INP lookback: 0 0
RRC bits: 0 0

Counters
Bearer 0
OHF: 0 30313053
OHFErr: 0 11
RS: 4085544128 1545845797
RSCorr: 9474905 145
RSUnCorr: 0 0
Bearer 1
OHF: 14677167 0
OHFErr: 0 0
RS: 146771050 0
RSCorr: 416 0
RSUnCorr: 0 0

Retransmit Counters
rtx_tx: 30626 0
rtx_c: 29262 0
rtx_uc: 0 0

G.INP Counters
LEFTRS: 7 0
minEFTR: 69321 0
errFreeBits: 862728315 0

Bearer 0
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 1361354975 0
Data Cells: 186524383 0
Drop Cells: 0
Bit Errors: 0 0

Bearer 1
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 0 0
Data Cells: 0 0
Drop Cells: 0
Bit Errors: 0 0

ES: 29 25
SES: 12 0
UAS: 64 53
AS: 235756

Bearer 0
INP: 50.00 0.00
INPRein: 1.00 0.00
delay: 0 0
PER: 0.00 7.80
OR: 0.01 32.78
AgR: 69389.98 12436.11

Bearer 1
INP: 4.50 0.00
INPRein: 4.50 0.00
delay: 3 0
PER: 16.06 0.01
OR: 79.68 0.01
AgR: 79.68 0.01

Bitswap: 171637/180722 1429/1435

I'm puzzled. Could some sort of trial be going on? Huawei cabinet, Billion 8800NL R2 modem/router.

Edited by RobertoS (Thu 16-Aug-18 00:26:06)

I see you are on Retransmittion High (IP Profile 91-92%).

I can tell this from Bearer 0 = INPRein 1

Retransmittion Low is Bearer 0 INPRein 0 (IP Profile 96.69%)

This thread might interest you.

Applying a moderate cap for a few days gives DLM a "nudge" to try Retx Low.

For some reason DLM prefers Retx High with Broadcom modems on a lower SNRM target. Very very few have retx low.

From that thread you can see 3 out of 3 of us have so far been successful in getting a nice little throughput boost.

Doubt it is some sort of trial. Normal SNRM variation looks like, although the quick resyncs could be a test of SRA that BT was testing at some stage, is SRA on on the router?

Try the Zyxel router, pretty stable SNRM on my line.

I'm happy with the stability Tim. The thread wasn't about that. It was definitely sync'ing at 3.5dB each time, with the expected variance of 0.1 or 0.2dB by the time you can take a reading. (It's a long time since I had a modem that logged the figure used). Something else was also being changed though as the sync increased each time by significant amounts. This is the sort of situation where MDWS was brilliant.

There was a final one a few days ago at 3dB with the speeds shown in my sig and it looks to be rock solid.

My ZyXel is out of the question because I wouldn't trust it with a factory downloaded update that may or may not cure the IPv6 problem that caused me to ditch it in the first place.

Neither do I seem to be able to use it as a modem to my currently unused ASUS RT-N66U. All the documentation about using it in bridge mode refers to it putting in router-only mode with a modem feeding it. The opposite of what is required. (Which is how AA have got my settings, which I have changed several times then put it on a shelf and bought the Billion).

I do keep wondering about replacing it with something with the previous (better) Broadcom chipset. The general view here seems to be a few more Mbps might be possible, but 70/14 is fine really. I may as well save the money.

If you could find the Billion 8800 V1 on ebay that has the same chipset as the Zyxel, I found putting the Zyxel into bridge mode easy with the help from a Zen page.

https://support.zen.co.uk/kb/Knowledgebase/ZyXEL-VMG...

At 18:16 Thursday 23rd I had another re-sync . Previous sync was 70615/12459 as reported here:-

In reply to a post by RobertoS:

This is the sort of situation where MDWS was brilliant.

There was a final one a few days ago at 3dB with the speeds shown in my sig and it looks to be rock solid.

That was in fact at about 05:10 on Friday 17. This time it was once again at 3dB, 70745/12295.

xdslctl info --show
xdslctl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 12295 Kbps, Downstream rate = 71620 Kbps
Bearer: 0, Upstream rate = 12295 Kbps, Downstream rate = 70745 Kbps
Bearer: 1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 3.0 6.0
Attn(dB): 19.4 0.0
Pwr(dBm): 13.7 7.6

VDSL2 framing
Bearer 0
MSGc: -6 26
B: 243 237
M: 1 1
T: 0 50
R: 10 16
S: 0.1098 0.6154
L: 18502 3302
D: 8 1
I: 254 127
N: 254 254
Q: 8 0
V: 0 0
RxQueue: 80 0
TxQueue: 20 0
G.INP Framing: 18 0
G.INP lookback: 20 0
RRC bits: 0 24
Bearer 1
MSGc: 154 -6
B: 0 0
M: 2 0
T: 2 0
R: 16 0
S: 6.4000 0.0000
L: 40 0
D: 3 0
I: 32 0
N: 32 0
Q: 0 0
V: 0 0
RxQueue: 0 0
TxQueue: 0 0
G.INP Framing: 0 0
G.INP lookback: 0 0
RRC bits: 0 0

Counters
Bearer 0
OHF: 0 3127835
OHFErr: 0 2
RS: 872740680 156373431
RSCorr: 20672060 12
RSUnCorr: 0 0
Bearer 1
OHF: 1497711 0
OHFErr: 0 0
RS: 14976498 0
RSCorr: 34000 0
RSUnCorr: 0 0

Retransmit Counters
rtx_tx: 120186 0
rtx_c: 118458 0
rtx_uc: 0 0

G.INP Counters
LEFTRS: 0 0
minEFTR: 70740 0
errFreeBits: 638317407 0

Bearer 0
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 3269303667 0
Data Cells: 8942920 0
Drop Cells: 0
Bit Errors: 0 0

Bearer 1
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 0 0
Data Cells: 0 0
Drop Cells: 0
Bit Errors: 0 0

ES: 160 27
SES: 21 0
UAS: 107 86
AS: 24057

Bearer 0
INP: 51.00 0.00
INPRein: 1.00 0.00
delay: 0 0
PER: 0.00 7.72
OR: 0.01 33.15
AgR: 70817.66 12327.84

Bearer 1
INP: 4.50 0.00
INPRein: 4.50 0.00
delay: 3 0
PER: 16.06 0.01
OR: 79.68 0.01
AgR: 79.68 0.01

Bitswap: 18641/19842 484/486

I haven't checked the two sets of stats yet to see if any interesting differences are present. Bog-eyed and off to bed.

Looks like DLM has put you on a 3db profile. Be fine as long as it is stable, but expect some variation of SNRM.

It did that on the 17th, like I said .

I still believe I had 3 syncs at 3.5dB. Attainable is gradually creeping up again. Up and down of course with the noise, but bit swapping or something trends it upwards and after a while is likely to re-sync again. Like it just did.