Kitz Forum

Broadband Related => ADSL Issues => Topic started by: Weaver on June 18, 2022, 04:49:10 PM

Title: Initial upstream SNRM too low, post retrain / resync
Post by: Weaver on June 18, 2022, 04:49:10 PM
My link #1 resynced / retrained this morning and came back up the wrong initial upstream SNRM. The upstream target is supposed to be 6 dB, but the new initial upstream SNRM was only ~3.7 dB. Later, when, by pure chance, I happened to run my DSL health check program, it rang an alarm bell and reported the upstream SNRM value as being too low compared to the expected value. I forced a resync and that fixed the problem.

Has anyone else ever seen this kind of problem?

The problem resync this morning was because of a massive interference blip; a large burst of downstream CRCs was seen on all links - link 1, 2 and link 4 - but for some reason, the burst problem was much worse on link 1, so only that link dropped. The downstream CRC burst count in the surrounding 30s interval was, for each DSL link:
   link 1: 660 CRCs,
   link 2:  20 CRCs,
   link 4:  11 CRCs.
Interference pick-up on the downstream side of link 1 is also visible in the CRC spike in the graphs below and also where the link 1 downstream SNRM drops quite low; from something near the 3.0 dB downstream target down to ~0.8dB.

Background: This problem has not been seen before in the few years since I first became able to monitor the modems’ stats easily. This is BT 21CN ADSL2. By the way, using Andrews and Arnold’s control server, I can set the upstream target myself to either 6dB or 9dB. I have PhyR (like G.INP) for downstream, but not upstream unfortunately. Even that did not manage to successfully defend against this burst, success presumably depending on the burst duration.

This incorrect initial SNRM was really bad news as it would have meant weeks of unreliable upstream because such a low upstream SNRM does mean a high CRC error count; upstream 6dB really is a minimum for reliable operation.

I wonder if a very large interleave would cure this kind of burst, but perhaps it was way too long in duration anyway. Perhaps the duration drives the decision to drop the link or keep going and just wait for it to pass. I don’t understand the decision logic. That logic was certainly wrong and resulted in a state that was an absolute mess.

Graphs - SNRM vs time, green is upstream:

(https://i.postimg.cc/YCsXJQjx/DEB8-D3-F3-5362-4-C98-A75-A-D71-C7-A460-B1-C.png)



Graph of CRCs vs time:

(https://i.postimg.cc/13wnQQ3Y/D6-B86-E31-42-CF-46-D6-A44-F-D6-B932568-B0-D.png)



Graph of FECs vs time:

(https://i.postimg.cc/wBNS2wFN/C7-FEC7-C8-912-E-4-F62-A9-E0-35-B4-B8-E9758-A.png)



Current detailed stats, as at 2022-06-18 15:48 UTC, well after problem fixed, unfortunately  :

Code: [Select]
xdslctl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 8000
Last initialization procedure status: 0
Max: Upstream rate = 570 Kbps, Downstream rate = 3056 Kbps
Bearer: 0, Upstream rate = 646 Kbps, Downstream rate = 2755 Kbps

Link Power State: L0
Mode: ADSL2 Annex A
TPS-TC: ATM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 2.9 5.8
Attn(dB): 65.0 41.0
Pwr(dBm): 18.2 12.4

ADSL2 framing
Bearer 0
MSGc: 53 11
B: 18 78
M: 8 1
T: 5 1
R: 16 16
S: 1.7364 3.8579
L: 774 197
D: 1 8

Counters
Bearer 0
SF: 1245508 200199
SFErr: 3 20
RS: 45928127 3601664
RSCorr: 7242 10092
RSUnCorr: 3 0

ReXmt: 1193 0
ReXmtCorr: 1193 0
ReXmtUnCorr: 3 0

Bearer 0
HEC: 2 24
OCD: 0 0
LCD: 0 0
Total Cells: 130331790 30600462
Data Cells: 5077585 1722635
Drop Cells: 0
Bit Errors: 180 2776

ES: 1403 3039
SES: 162 6
UAS: 1502 1364
AS: 20061

Bearer 0
INP: 29.00 2.50
INPRein: 0.00 0.00
delay: 8 8
PER: 16.00 16.39
OR: 29.48 8.29
AgR: 2771.88 652.73

Bitswap: 3333/3333 15/15

Total time = 52 days 12 hours 5 min 58 sec
FEC: 4159597 3118687
CRC: 9669 4479
ES: 1403 3039
SES: 162 6
UAS: 1502 1364
LOS: 8 0
LOF: 62 0
LOM: 0 0
Latest 15 minutes time = 5 min 58 sec
FEC: 78 283
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 173 765
CRC: 0 4
ES: 0 3
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 12 hours 5 min 58 sec
FEC: 23611 20193
CRC: 1196 43
ES: 75 28
SES: 23 0
UAS: 125 104
LOS: 1 0
LOF: 9 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 29570 26828
CRC: 283 171
ES: 117 123
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Since Link time = 5 hours 34 min 19 sec
FEC: 7242 10092
CRC: 3 20
ES: 3 15
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
NTR: mipsCntAtNtr=0 ncoCntAtNtr=0