I Need Help - Don't Faint!

[Ignitionnet]

I'm at a loss as to what could be wrong with my cable.

I have had upstream issues of varying degrees for a while which seemed largely solved by some work on the network, I now seem to have a new problem.

Thu Sep 17 03:24:23 2009 Thu Sep 17 03:24:23 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync;

Thu Sep 17 08:16:04 2009 Thu Sep 17 08:16:04 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync

Thu Sep 17 08:42:32 2009 Thu Sep 17 08:42:32 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync

Which happens, however I was able to lock onto the DOCSIS 1 network no problem. Modem synced up with that then was kicked over to the DOCSIS 3 network where it failed to acquire the downstream and started ranging again.

And off we go again today:

Modem lost sync at 3:20:

Mon Sep 21 14:20:12 2009 Mon Sep 21 14:20:12 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;

It could not re-acquire downstream for a fair while:

Mon Sep 21 14:20:40 2009 Mon Sep 21 14:20:40 2009 Critical (3) Received Response to Broadcast Maintenance Request, But no Unicast Maintenance opportunities received - T4 time out

Mon Sep 21 14:21:53 2009 Mon Sep 21 14:21:53 2009 Warning (5) MDD message timeout;CMTS-MAC=00:00:00:00:00:00;CM-QOS=1.1;CM-VER=3.0;

Mon Sep 21 14:24:51 2009 Mon Sep 21 14:24:51 2009 Warning (5) MDD message timeout;CMTS-MAC=00:00:00:00:00:00;CM-QOS=1.1;CM-VER=3.0;

and eventually after bouncing between the DOCSIS 1.0 network and failing to re-acquire the DOCSIS 3 one came back online 30 minutes later:

Mon Sep 21 14:52:17 2009 Mon Sep 21 14:52:17 2009 Notice (6) TLV-11 - unrecognized OID;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;

I can't explain these issues, all power levels are absolutely fine:

Frequency 299000000 307000000 315000000 323000000
Lock Status
(QAM Lock/FEC Sync/MPEG Lock) Y/Y/Y Y/Y/Y Y/Y/Y Y/Y/Y
Channel Id 10 11 12 13
Modulation 256QAM 256QAM 256QAM 256QAM
Symbol Rate
(Msym/sec) 6.952 6.952 6.952 6.952
Power Level
(dBmV) 4.45 4.65 4.78 5.14
RxMER
(dB) 38.26 38.61 38.61 39.40
Correctable
Codewords 1 5 4 2
Uncorrectable
Codewords 289 286 286 311

US-1
Channel Type 1.0
Channel ID 1
Frequency
(Hz) 25800000
Ranging Status Success
Modulation 16QAM
Symbol Rate
(KSym/sec) 2560
Mini-Slot Size 2
Power Level
(dBmV) 41.00
T1 Timeouts 0
T2 Timeouts 0
T3 Timeouts 197
T4 Timeouts 10

Any suggestions welcome. Previous work on the node did remove the T3s for a couple of weeks then they started to reoccur. I have not noticed any performance issues from this however.

[xocemp]

Any chance of another snapshot of the codewords please BB?

[Ignitionnet]

Not incrementing I'm afraid, so if there's bursts of downstream noise they are very intermittent. Also only seems to affect the D3 frequencies, modem was acquiring 331MHz np, then it received its' DCC from the VXR... and failed to sync at 299.

RxMER
(dB) 38.26 38.61 38.61 39.40
Correctable
Codewords 1 6 4 2
Uncorrectable
Codewords 289 286 286 311

[Kymmy]

Hmmm, will you not post your MAC address :mad:

I wonder how many cloners are now trying 00:00:00:00:00:00 on thier modem :p:

[Ignitionnet]

Hmmm, will you not post your MAC address :mad:

I wonder how many cloners are now trying 00:00:00:00:00:00 on thier modem :p:

OK I won't post it.

Any cloners who didn't read the TS in CMTS-MAC might have a go :)

It was important as part of the error message, pointed out that the modem didn't have a MAC for the CMTS, as it never saw an MDD.

[xocemp]

BB ask one of the nice chaps on the groups to check your flaps & P-Adj.

Non of this makes sense, from what you describe then I'd have pointed the finger at amps are filtered to stop frequencies above 40Mhz then modems are not going to lock. Though this is would be more likely to cause you to not lock to the D3.

[Ignitionnet]

BB ask one of the nice chaps on the groups to check your flaps & P-Adj.

Non of this makes sense, from what you describe then I'd have pointed the finger at amps are filtered to stop frequencies above 40Mhz then modems are not going to lock. Though this is would be more likely to cause you to not lock to the D3.

Yeah the amps start to roll off at 40MHz, the subsplit is at 42MHz so this is expected.

When DOCSIS 3 came to town they took an upstream frequency from the DOCSIS 1 network so evidently there were capacity restraints on the network.

[xocemp]

Flaps & P-Adj
Were they forthcoming?

[Ignitionnet]

They want to send a service tech, but I've been down that road more than once. I'm utterly convinced a service tech will only delay things as I'll have to get them then it gets pushed to networks.

There's a problem at Mortlake hub, or shonky optics, or a shonky amp on the cascade.

I've asked for output of:

sh cab mod ver
sh cab mod phy
sh cab flap-list | i xxxx.xxxx.xxxx

[xocemp]

Agreed a service tech is not going to help. I'm inclined to go with the 10 year old amp, who'd have though that 10 year old kit would cause problems :rolleyes:

Have one of the chaps check for discards on your cable.

show int cable x/y downstream /x/y/z

[Ignitionnet]

Agreed a service tech is not going to help. I'm inclined to go with the 10 year old amp, who'd have though that 10 year old kit would cause problems :rolleyes:

Have one of the chaps check for discards on your cable.

show int cable x/y downstream /x/y/z

Only 10 years old? You're having a laugh :D

*Boing*

Mon Sep 21 20:10:46 2009 Mon Sep 21 20:10:46 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;
Mon Sep 21 20:11:15 2009 Mon Sep 21 20:11:15 2009 Critical (3) Received Response to Broadcast Maintenance Request, But no Unicast Maintenance opportunities received - T4 time out;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;
Mon Sep 21 20:12:27 2009 Mon Sep 21 20:15:15 2009 Warning (5) MDD message timeout;CMTS-MAC=00:00:00:00:00:00;CM-QOS=1.1;CM-VER=3.0;
Mon Sep 21 20:15:43 2009 Mon Sep 21 20:15:43 2009 Notice (6) TLV-11 - unrecognized OID;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;

EDIT:

Mon Sep 21 20:21:38 2009 Mon Sep 21 20:21:38 2009 Critical (3) SYNC Timing Synchronization failure - Loss of Sync;CMTS-MAC=00:30:b8:d0:ac:50;CM-QOS=1.1;CM-VER=3.0;

EDIT 2:

Is it normal for downstream IDs to change?

Previously it was:

Frequency 299000000 307000000 315000000 323000000
Lock Status
(QAM Lock/FEC Sync/MPEG Lock) Y/Y/Y Y/Y/Y Y/Y/Y Y/Y/Y
Channel Id 10 11 12 13

Now it's:

DS-1 DS-2 DS-3 DS-4
Frequency 299000000 307000000 315000000 323000000
Lock Status
(QAM Lock/FEC Sync/MPEG Lock) Y/Y/Y Y/Y/Y Y/Y/Y Y/Y/Y
Channel Id 74 75 76 77

Those IDs come from the MDD / DDS messages, are they just cosmetic on the VM network?

[Welshchris]

Broadbandings this is the problem i was getting weeks ago and i was told it was because my upstream power level was hitting 59dBmv and was dropping and attempting to lock onto a different channel.

[Ignitionnet]

Thanks, sadly if it were that easy I wouldn't be posting here :)

Power Level (dBmV) 41.00

As per the above errors the issues are downstream related. I'm not seeing a burst of T3s followed by a T4 forcing a resync I'm seeing loss downstream firstly.

EDIT: Oh for 10 minutes of enable access on my CMTS... :(

[xocemp]

There are typically 299, 307 & 315, 298.5, 306.5 & 314.5 323.5

Stop getting T4's damn it!
So your not getting the "keep alive" requests from the CMTS for this to occur, this can't be because of shonky upstream / lack of bandwidth.

xoce-edit Ok, read your above post. Cool, we can keep on looking at D/S.

xoce-edit2 Channel Id's not freq's, sorry dude its been a long day.
Your on a BSR so I'm going to come clean and say I have no idea with these beasts, though I would imagine that the configuration has been set for the frequencies to be sent / broadcast rather than the Channel Id's.

xoce-edit3 Found some documentation.
http://www.cisco.com/en/US/docs/ios/12_0t/12_0t4/feature/guide/downchan.html

[Ignitionnet]

Nah the T3s are shonky upstream, T4s are a symptom of the shonky downstream :D

*Sigh* it's goosed. Shame Telewest ran out of dosh buying this area and couldn't afford to sort it out.

I could do with a spectrum analyser, anyone know how to reprogram an Ambit 256 into one?

EDIT: Cheers for that dude, as per my previous update:



Those IDs come from the MDD / DDS messages, are they just cosmetic on the VM network?

In the same messages that carry the frequencies, modulations and channel widths of the downstreams are IDs, it's an 8 bit field. You can find more on this from the standards at http://www.scte.org

It would seem a bit odd for these IDs to change but far from impossible.

http://www.scte.org/documents/pdf/Standards/Top%20Ten/ANSI_SCTE%20135-2%202008.pdf

6.4.28.1.1 Downstream Active Channel List TLV
Each instance of this TLV represents one downstream channel in the MAC Domain. The CMTS MAY include this
TLV more than once in a given MDD message.
When sending this message on a primary-capable downstream channel, the CMTS MUST include a Downstream
Active Channel List TLV for every downstream channel in every MD-DS-SG that contains the current channel.

Type Length Value
1.1 1 Channel ID: 1 byte. The Downstream Channel ID of the channel being listed.
1.2 4 Frequency: 4 bytes. The center frequency of the downstream channel (Hz).
This TLV is intended only to assist CMs in speeding the acquisition of new channels
prior to the completion of registration.

---------- Post added at 22:50 ---------- Previous post was at 21:56 ----------

Upstream stats (which don't get reset upon loss of sync) now:

Power Level (dBmV) 40.00
T1 Timeouts 0
T2 Timeouts 0
T3 Timeouts 201
T4 Timeouts 12

3 hours ago:

Power Level (dBmV) 41.00
T1 Timeouts 0
T2 Timeouts 0
T3 Timeouts 197
T4 Timeouts 10

Power Level
(dBmV) 39.50

[Ignitionnet]

And dead again. Using a neighbour's wireless. I guess you guys will be free of me for a while while I direct my wrath to Virgin for the 3rd upstream / return path network outage since last May, ignoring the degraded service issues last year in between.

*Goes off to attack fibre node with a spade*

[Ignitionnet]

Just to complete called technical support and had a quick chat, upstream is hosed, no modems online, poor SNR on the node which might well be because there's nothing online :)

I posted to the support newsgroup exactly a week ago mentioning that I'd seen some misbehaviour on the upstream - transmit power had a swing from 36dBmV to 51.75dBmV almost instantly then across the course of the week down to 46.75dBmV before dying earlier today.

Modem is jammed at 58.21dBmV transmit power, T3 timeouts are counting up.

[Mr Angry]

Just to complete called technical support and had a quick chat, upstream is hosed, no modems online, poor SNR on the node which might well be because there's nothing online :)

I posted to the support newsgroup exactly a week ago mentioning that I'd seen some misbehaviour on the upstream - transmit power had a swing from 36dBmV to 51.75dBmV almost instantly then across the course of the week down to 46.75dBmV before dying earlier today.

Modem is jammed at 58.21dBmV transmit power, T3 timeouts are counting up.


Does your chewing gum lose its flavour on the bedpost overnight?