Need help with ongoing packet loss issue

[Effayy]

My Setup

Running a CODA-4582U in bridge mode, connected to an ASUS router providing both wired/wireless connectivity to home devices.  Both my home and work PCs are wired, as are my PS5 and XBox One.  The rest of the family (wife and 4 kids) each have their respective phones and other gadgets all wirelessly connected to the ASUS router.

I live in a townhome in a reasonably new neighbourhood (our street is only 11 years old). 

Symptoms

For the past few months we've experienced packet loss which causes interruptions in all our activities. Work VDI will disconnect, Zoom/Teams meetings will stall out or pause temporarily. Online gaming will lose connection to host servers, etc.   Rebooting the cablemodem will calm things down for anywhere between 12 hours and 3 days, Once the packet loss starts up again, it will continue regardless of time of day until the cablemodem is rebooted again. 

Investigation

At first I thought it was a hardware issue, so had my original CODA-4582 switched out for a refurb CODA-4582U.  This did not resolve the issue.

When the packet loss started acting up, before rebooting I ran continual ping tests for an hour from my wired PC against the following:

• cablemodem (192.168.100.1):  Rock solid. 0 packet loss. 
  • This means it is not a problem with my internal network, correct?
• CMTS (174.x.x.1):  22 timeouts in 1 hour. (averaged roughly once every couple of minutes)
  • This IP was the 2nd hop in my traceroute so I assume it's CMTS?
• DNS (64.71.255.204):  25 timeouts in 1 hour.

I also tested restarting the ASUS router without rebooting the cablemodem, and the packet loss persisted through the reboot.

During a period of packet loss, I spoke with support over the phone who agreed that there was likely an issue somewhere between my house and the CMTS.  He remotely rebooted my modem, we both agreed that the issue would likely occur again anyway so he scheduled a tech to come out.

Tech came yesterday and checked the lines outside and told me everything was fine. He said that the remote reboot was different than what I would do on my end, and that it likely fixed the problem.  I reluctantly let him go but wasn't satisfied with his reasoning on this. Unfortunately when he was on-site the network was behaving itself so I didn't really have a leg to stand on at that point.  As expected however, 24 hours later the issue returned.

Based on scouring through threads in here over the past couple of weeks (thanks @Datalink you're a gold mine of information!), my theory is that it's either a cabling issue between my house and the CMTS, or the neighbourhood is using more bandwidth than the current infrastructure can hold.  I asked the phone rep if he was able to determine the neighbourhood bandwidth saturation from his end and he seemed to avoid answering the question.

Is there anything else at this point that I can try to help shed more light on the root problem?  I'm somewhat technical but admittedly not familiar with cable networking outside of what I've read on these forums.  

Below are my DOCSIS stats.  I believe they're within normal range.

Thanks in advance!

Downstream Overview

Port ID	Frequency (MHz)	Modulation	Signal strength (dBmV)	Channel ID	Signal noise ratio (dB)
1	615000000	QAM256	-1.500	11	37.355
2	591000000	QAM256	-0.799	7	37.355
3	597000000	QAM256	-0.799	8	37.636
4	603000000	QAM256	-1.099	9	37.355
5	849000000	QAM256	-3.700	2	36.386
6	855000000	QAM256	-4.000	3	36.386
7	861000000	QAM256	-4.500	4	36.386
8	579000000	QAM256	-0.400	5	37.355
9	585000000	QAM256	-0.500	6	37.355
10	609000000	QAM256	-1.299	10	37.355
11	279000000	QAM256	6.599	1	38.983
12	621000000	QAM256	-1.500	12	37.355
13	633000000	QAM256	-2.099	13	36.609
14	639000000	QAM256	-2.400	14	36.386
15	645000000	QAM256	-2.599	15	36.609
16	651000000	QAM256	-2.500	16	36.609
17	657000000	QAM256	-2.500	17	36.609
18	663000000	QAM256	-2.599	18	36.609
19	669000000	QAM256	-2.599	19	36.609
20	675000000	QAM256	-2.900	20	36.609
21	681000000	QAM256	-2.799	21	36.609
22	687000000	QAM256	-2.900	22	36.386
23	693000000	QAM256	-2.900	23	36.609
24	699000000	QAM256	-2.799	24	36.609
25	705000000	QAM256	-2.799	25	36.609
26	711000000	QAM256	-3.099	26	36.609
27	717000000	QAM256	-3.000	27	36.609
28	723000000	QAM256	-3.599	28	36.386
29	825000000	QAM256	-3.500	29	36.609
30	831000000	QAM256	-3.599	30	36.609
31	837000000	QAM256	-3.500	31	36.609
32	843000000	QAM256	-3.599	32	36.609

OFDM Downstream Overview

Receiver	FFT type	Subcarr 0 Frequency(MHz)	PLC locked	NCP locked	MDC1 locked	PLC power(dBmv)
0	NA	NA	NO	NO	NO	NA
1	4K	275600000	YES	YES	YES	5.599998

Upstream Overview

Port ID	Frequency (MHz)	Modulation	Signal strength (dBmV)	Channel ID	Bandwidth
1	38700000	64QAM	34.270	4	6400000
2	21100000	64QAM	33.260	1	3200000
3	32300000	64QAM	34.270	3	6400000
4	25900000	64QAM	34.270	2	6400000
5	0	QAM_NONE	-	---	1600000
6	0	QAM_NONE	-	---	1600000
7	0	QAM_NONE	-	---	1600000
8	0	QAM_NONE	-	---	1600000

OFDM/OFDMA Overview

Channel Index	State	lin Digital Att	Digital Att	BW (sc's*fft)	Report Power	Report Power1_6	FFT Size
0	DISABLED	0.0000	0.0000	0.0000	0.0000	0.0000	2K
1	DISABLED	0.0000	0.0000	0.0000	0.0000	0.0000	2K

**Labels Added**

[Datalink]

@Effayy its a little late at the moment, so I'll keep this short.  You have an external cable and/or connector problem.  The fact that the tech rebooted the modem, and voila, it worked for a period of time basically proves it.  The reboot will return the cable back to a normal operative state for a period of time.  That period of time will depend on how badly the cable and its connectors have degraded.  My bet is that the modem reboot returns the cable back to a normal capacitive state for a period of time.  After that, the cable returns to its failed state.  Each cable type has a capacitive value based on its construction.  A break in the cable or water ingress would alter that value and short the internal copper conductor to the external cable ground, causing a disconnect.  For whatever reason, a reboot resolves the situation for a period of time, but it doesn't fix the problem.  

Call tech support and register yet another complaint.  Call every day if you have to.  That will build a history of poor service over a short period of time. 

Don't let tech support or the field tech reboot the modem and declare victory.  That just kicks the can down the road and makes the tech look good on paper ..... another customer complaint resolved (not really).

Tech support will insist on another contractor tech visit. That's the routine unfortunately.  After that visit, you can ask for a Senior Tech (real Rogers tech).  I'd say that this is going to take more than one or two visits to resolve.  

Assuming that you end up with yet another tech, ask if he can run a Time domain reflectometer test on the external cable.  That sends a pulse down the cable that bounces back at the end of the cable, or at the location of a break.  If the cable is tested from both ends, then the timed response should be the same.  If there's a break in the cable or water ingress at some point, then the measured cable lengths would be different.  That's an obvious sign that the cable must be replaced.  I'm assuming that the townhome has all of the external cable enclosures at one end of the building.  So, there would be a cable run from your townhome to that external NID.  Then, from the NID, there would be another run, either underground or overhead, running to the local tap.  If that run is underground, then the tech can test that run using a TDR.  I'd bet that he could also run a TDR test from the external NID to the internal end of the cable, just to check for the possibility of an internal cable fault. 

So, that's the starting point, or so I think.  A closer examination of the cabling, from the townhome to the external NID, and from the NID to the local tap. 

Fwiw, if this was a very large townhome complex, there's always the possibility of an internal Multiple Dwelling Unit (MDU) supplying data services to all of the units.  That's basically an internal local tap, designed for large buildings.  These are pretty reliable, but, if an when they go wrong, they're a pain to diagnose and fix.  I don't believe that contractor techs are allowed to do anything with an MDU other than connect and disconnect.  Anything other than that would require a Rogers tech.  

So, one question for the next tech is "how is the townhome connected?" Though an external NID and then out to a local tap, or thru an internal MDU?  The answer to that will also determine what needs to be done in the future.

Ask your neighbours if they're experiencing any disconnects.  If they are, that points the finger at the local tap or beyond, or at the MDU, if there is one located in the building complex. 

Here's what the TDRs look like:

https://megger.com/products/cable-fault-test-and-diagnostics/cable-fault-locating-equipment/time-dom...

[hoboryan]

Have you checked through the speedtest app (iphone/android)? How much % is your packet loss?

[Effayy]

Thanks for the awesome advice @Datalink. 

I started seeing packet loss again tonight, and made another call in to support.  I essentially said that I know this will keep happening, and the next time we schedule a tech call I would prefer it to be a senior tech as I feel the issue lies in a cabling malfunction somewhere between my house and the NID, or beyond.  She agreed that rebooting was just a band-aid and put a note on file.

She set up a monitoring ticket and when I start to see packet loss in the next couple of days (which based on history is inevitable), I'll be able to call back and (fingers crossed) get that senior tech to come out. 

My townhome isn't a large complex. just a row of 4 houses, and I'm the end-unit.  Likely the infrastructure is the home --> NID --> tap scenario you mentioned.  Hopefully that makes like a little easier for the tech to determine root cause.

I'll post back here as things progress.  Hoping for a happy ending.

[Effayy]

My speedtest app doesn't show % packet loss (or jitter either).  All tests I've done have been from my PC so that I can take any potential WiFi problem out of the equation.  I have PingPlotter installed as well, so perhaps I can get some numbers from that.

[hoboryan]

You can try the IOS app or the windows desktop app (if you drill into details it will show packet loss). I'm having similar problems to you and this is what they told me today (posting from another thread).

So called again today and they escalated to a supervisor. Who pinged my modem and confirmed 10% packet loss (not sure if he was just making it up). He also pinged my neighbours and confirmed they had no packet loss.

They're sending a guy to put in a temporary cable from the curb under the assumption that the cable leading to my house is degraded. Who knows if it'll work.