CM3000 – repeated T3 timeouts, SYNC failures, intermittent disconnects

Hi Abby, thanks for the reply.

Timing: It's not limited to one part of the day — I get it throughout. The download latency is fairly consistent whenever I'm active (elevated ping with spikes to 500ms+), and the disconnects can happen anytime, including while gaming in the evening. The T3 timeouts in the modem log do cluster somewhat in the afternoon/evening peak window, but they also occur at other times, including overnight. So there are really two things happening: constant latency/instability, plus clustered T3/SYNC events on top of it.

Troubleshooting already completed (this has been exhaustive on my end):

• Power-cycled the modem many times — both via the Xfinity app and physical unplugs
• Restarted and reconfigured the router multiple times
• Verified the coax path myself: single ~10 ft drop from the side of the house to the modem, no splitters, connectors hand-tight
• Had a technician on-site — he metered signal at the tap/street and at the house and confirmed the outside plant and drop are clean; the problem has persisted since his visit
• Captured the full modem signal tables (downstream + upstream) and event logs. Note: the modem only retains about 5 weeks of event history, but this problem long predates that — I've been dealing with it far longer; 5 weeks is simply as far back as the log buffer goes
• Confirmed downstream is pristine (31 channels locked, SNR 40–42 dB, power in spec, zero uncorrectables), which isolates the problem to the upstream/return path
• Identified my firmware as V6.01.04 (the successor to the pulled V6.01.03) and am aware of its reported upstream issues
• Per the tech's suggestion, I'm eliminating any possible premises interference (dedicated clean power, separating and relocating equipment/cabling) — but the T3 timeouts and SYNC failures are DOCSIS-layer upstream events that occur independent of my LAN setup

Key evidence: In just the ~5 weeks the modem's log even retains, there are 27+ Critical T3 timeouts and the frequency is increasing — including "16 consecutive T3 timeouts while trying to range on upstream channel 1 – retries exhausted," plus repeated SYNC timing failures. This has been ongoing much longer; the log only captures the most recent window. Upstream transmit power is also elevated at ~43 dBmV on a ~10 ft drop with no splitters.

Given that reboots, a tech visit, and premises remediation haven't resolved it, could you check my upstream/return-path signal and SNR at the CMTS, look for ingress/noise on my node, and verify my firmware? I'd like this escalated to maintenance / advanced repair rather than another basic tech visit. Happy to share account details by private message.

I'm still having issues. Tech said everything was fine. No idea what to do next. If I find out anything I promise to come back and update. At this point I need to gather more info.

Same on your end please... let me know what the tech visit reveals. Ill post back if I get anything. 

Nebby5k when is your appointment scheduled for? 

@XfinityJosephA​ It’s scheduled for this Monday

@XfinityAbby @XfinityEricB @XfinityJeff Can anyone from Xfinity help with this? 

@user_cb0tlj​ I'm monitoring this with bated breath, same issue.

Been running for years without any issues, sometimes last month there was "maintenance" on our network that resulted in constant issues. 

Had a tech out, everything checked out fine. At first it was maybe 20-30 seconds a week, last week it escalated to being down more often than not. Been up for less than 2 hours today, and all the logs say it T3/T4 timeouts with a B-INIT-RNG Failure.

Yes and I have done that numerous times. How many times do we have to keep going over the same information and then have someone ask me if I restarted my modem? Or we have the same tech who has come out multiple times to check the local line and tell me he can't find the problem. We are way beyond the basics here. Xfinity has all of the information. You guys need to escalate this. 

@user_cb0tlj We will need to look into the connection and start the process of getting this looked into. It may involve tech visits before an escalation can take place.