[169337] in North American Network Operators' Group
Re: comcast business service
daemon@ATHENA.MIT.EDU (shawn wilson)
Fri Feb 21 05:23:54 2014
In-Reply-To: <FA7CAADC-FB8A-47C9-927F-7BE9A1A21C06@iwiring.net>
From: shawn wilson <ag4ve.us@gmail.com>
Date: Fri, 21 Feb 2014 05:23:07 -0500
To: Dan Shoop <shoop@iwiring.net>
Cc: Nanog <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Works:
Downstream Channel
Downstream Frequency525000000 Hz561000000 Hz567000000 Hz573000000 Hz5790000=
00 Hz
Lock StatusLockedLockedLockedLockedLocked
Modulation256 QAM256 QAM256 QAM256 QAM256 QAM
Symbol Rate5.360537 Msym/sec5.360537 Msym/sec5.360537 Msym/sec5.360537
Msym/sec5.360537 Msym/sec
Downstream Power 2.2 dBmV 3.8 dBmV 3.0 dBmV 2.9 dBmV 2.9 dBmV
SNR41.2 dBmV40.8 dBmV40.5 dBmV40.9 dBmV41.0 dBmV
Upstream Channel
Upstream Frequency36000000 Hz29400000 Hz22800000 Hz0 Hz
Lock StatusLockedLockedLockedNot Locked
ModulationATDMAATDMAATDMAUnknown
Symbol Rate5120 sym/sec5120 sym/sec5120 sym/sec0 sym/sec
Upstream Power46.2 dBmV46.2 dBmV46.2 dBmV0 dBmV
--- 8.8.8.8 ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9013ms
rtt min/avg/max/mdev =3D 23.066/27.049/35.627/4.825 ms
Not working:
Downstream Channel
Downstream Frequency525000000 Hz561000000 Hz567000000 Hz573000000 Hz5790000=
00 Hz
Lock StatusLockedLockedLockedLockedLocked
Modulation256 QAM256 QAM256 QAM256 QAM256 QAM
Symbol Rate5.360537 Msym/sec5.360537 Msym/sec5.360537 Msym/sec5.360537
Msym/sec5.360537 Msym/sec
Downstream Power 2.2 dBmV 3.8 dBmV 2.9 dBmV 2.8 dBmV 2.9 dBmV
SNR41.4 dBmV40.8 dBmV40.4 dBmV41.0 dBmV41.3 dBmV
Upstream Channel
Upstream Frequency36000000 Hz29400000 Hz22800000 Hz0 Hz
Lock StatusLockedLockedLockedNot Locked
ModulationATDMAATDMAATDMAUnknown
Symbol Rate5120 sym/sec5120 sym/sec5120 sym/sec0 sym/sec
Upstream Power46.5 dBmV46.5 dBmV46.5 dBmV0 dBmV
--- 8.8.8.8 ping statistics ---
233 packets transmitted, 232 received, 0% packet loss, time 232884ms
rtt min/avg/max/mdev =3D 23.431/1918.702/8758.161/2017.033 ms, pipe 9
I'm not seeing any big difference in SNR (and only slight differences
in upstream power) and everything else seems to be the same. Though,
since db is logarithmic, .3 might be enough to matter?
On Thu, Feb 20, 2014 at 4:14 PM, Dan Shoop <shoop@iwiring.net> wrote:
>
> On Feb 20, 2014, at 4:08 AM, shawn wilson <ag4ve.us@gmail.com> wrote:
>
>> A while ago I got Comcast's business service. Semi-idle connections
>> are get dropped (I haven't really diagnosed this - I just no that it
>> isn't the client or server but some network in between). However the
>> second and most obvious issue is that intermittently, the service will
>> grind to a halt:
>> --- 8.8.8.8 ping statistics ---
>> 37 packets transmitted, 34 received, 8% packet loss, time 36263ms
>> rtt min/avg/max/mdev =3D 398.821/5989.160/14407.055/3808.068 ms, pipe 15
>>
>> After a modem reboot, it goes normal:
>> --- 8.8.8.8 ping statistics ---
>> 4 packets transmitted, 4 received, 0% packet loss, time 3003ms
>> rtt min/avg/max/mdev =3D 23.181/23.920/24.298/0.474 ms
>>
>> This seems to happen about once or twice a day. I can't attribute it
>> to any type of traffic or number of connections. All of the rest of
>> the network equipment is the same and the behavior persists when a
>> computer is plugged directly into the modem. I called Comcast and they
>> said they didn't see anything even when I was experiencing ridiculous
>> ping times. I tend to think it's an issue with the 'modem' but I'm not
>> sure what the issue might be or how to reproduce it when asked to if I
>> tell them to look at it.
>
> I=E2=80=99ve seen this happen before with various cable ISPs. I=E2=80=99d=
concur with the poster suggesting intermittent noise on the cable segment =
as a likely culprit. Also if you have a cable modem that binds multiple cha=
nnels for higher bandwidth this can also be problematic, especially with th=
e noise. Signals will look good to the NOC but it=E2=80=99s not the signal =
=E2=80=9Clevel" that=E2=80=99s the issue it=E2=80=99s the signal to noise l=
evel. Noise has to be measured locally and techs don=E2=80=99t always check=
SNL.
>
> Also check to see if the packets aren=E2=80=99t actually being dropped bu=
t just taking longer than ping is looking for. Also check for out of sequen=
ce packets returned. These can indicate flapping of a bonded circuit or the=
bonded circuit experiencing noise. Try seeing if you disconnect everything=
and get a straight run to the demarc, with a know and tested out good cabl=
e, if the problem doesn=E2=80=99t ever occur. This could indicate noise on =
the cable in your premise. But I=E2=80=99ve experienced this same problem w=
ith noise coming through the demarc. I=E2=80=99ve also seen levels too hot =
beyond the demarc causing similar problems too.
>
> HTH.
>
>
> -d
>
> -----
>
> Dan Shoop
> shoop@iwiring.net
> 1-646-402-5293 (GoogleVoice)
>
>
>
>
