Didn't find what you were looking for?
We have advanced search options to make it easier to locate posts, questions and answers on this community.
More information can be found at Advanced Search Options
If you are looking for something specific, please check if someone else has already asked or answered the same question.
Is this normal? correctables/uncorrectables
I had a tech come out a few times net speed was dropping 300-400 less in the evenings with the 1 gig tier. Second one just decided to change the cable. Noted potentially with a small % that the cable was pinched a bit so went ahead. This is with 5hours + of uptime on the modem. I know the corrected on the OFDM is normal but is anything else unusual going on?
Modem is the Motorola MB8611
Answers
-
Hello @KERNOLSZ28;
Welcome to the Community Forums. I can understand the frustration of the internet not being stable. I reviewed the modem's history and it has been stable with the errors since your last service call. I did see a spike of errors prior to the last service call. It is possible that the last technician has fixed the problem that you have been having. If you continue to see problems with the internet then reach back out and we will check the equipment to see if the issue has returned.
0 -
Hi Edward_T,
Thank you for checking in on that. I stated both times the techs were here the speeds were dropping speed wise towards the evenings. 1 did find ingress on 2 lines not being used coming into the home so those were removed but that fixed issues those were causing not the one I was mentioning. So the last one did change the whole line from poll to modem. Im hoping that did the trick yes. I was just curious if the amount of errors is normal/nothing to worry about at all. esp on the OFDM because periodically while I looked at the modem before this last tech came out, the downstream icon would be flashing which means I guess it was bonding?
0 -
Hello @KERNOLSZ28;
The date you provided does not indicate a timeframe on when the errors occurred. You may see an error occasionally but does not necessarily indicate an issue. The issue is when the errors are high in a short period of time. Our systems record the errors in a historical timeframe. I do see a spike in high errors prior to the last service call. After the service call, the errors were at normal levels. The Downstream flashing can indicate that it was trying to lock onto a connection.
0 -
@Edward_T
My speeds are sitting at 875-900 and just checkin in on the modem and Ive had a massive spike on the OFDM channel. all else stayed the same. It didn't dissconnect during gaming but this is far different then yesterday as that was around 3.8k in 24 hours. This channel seems to be the one that has the issues.
0 -
Pushing the forward error correction envelope is literally a part of the equation in achieving the immense spectral density and, thus, astronomical speed of DOCSIS 3.1 OFDM and it looks like yours is doing the full-blown 192-MHz-bandwidth, possibly even with the top-end 4096-QAM constellation.
In any case, you'll be hard-pressed to find any authoritative assessment to suggest that there's anything bad about any of those numbers. Uncorrected rate less than 1% is supposedly fine, even for VoIP! More importantly, as long as it's less than 1%, then the corrected rate, even up to 100%, believe it or not, is rendered trivial and keep in mind how uncorrected codewords, for the most part, will end up being retransmitted in short order anyway.
You can't see the number of error-free codewords nor, thus, the total number of codewords but, in the theoretical albeit practically impossible worst-case scenario that none were error-free, with your OFDM channel, your uncorrected rate = uncorrected/(uncorrected + correctected) X 100 ≈ 0.0048%‼️☘️
🔗 ↳ … Again, let’s reinforce: in DOCSIS 3.1 it’s okay to have very high levels of correctable codeword errors as long as uncorrectable codewords are non-existent or less than 1%. …
P.S: With so many factors at play, there are, of course, no absolute guarantees but, if you can't ever break 1 Gbps via e.g. Ookla Speedtest even if just by a smidgeon, then i'd make sure that the modem, router and machine are all three at least 2.5-Gbps-Ethernet capable and, thus, that the cables are all least category 6 but that's just me.
0 -
Hi HT_Greenfield ,
Thanks for showing me that link, I will look into it. All of that would explain I never saw it while I was gaming/in discord with something streaming. My main concern in the first place was speed yeah. I get anywhere between 1050 and 1150 mostly during the day ( have a ROG GT AX6000 router ) that does have the 2.5gb ports and my pc is also 2.5gb. Its the evenings I see much lower. Before this last tech came out I was sitting around 600-700 during the evenings. The levels on the modem were more of a I wanted to make sure these were okay because The first tech did mentioned ingress on the lines that he removed and thought he saw it all gone. Mentioned possible over utilization and maintenance did come out and re-balance. I know its "up to" 1gig but When I was on the older 500 ultra + 100 boost I was 680 down every speed test no matter what. Never lost any bandwitdth. It seemed dropping almost 400-500 down was a little much with the gig tier. All during evenings only.
0 -
So based on your math, my channel 1 has a 31% uncorrectable rate. But the numbers are small in terms of frequent adding up. Does that mean the channel is having an issue? or no because the actual numbers are small in number.
0 -
Only if pretzel logic is your bag. Otherwise, again: You can't see the number of error-free codewords nor thus the total number of codewords so the logical question would be whether the total number of codewords in that channel could possibly be less than 100 X the number of its uncorrected codewords = 2400 in that channel in that same duration in which the OFDM channel literally had ~400,000 X as many correcteds alone.
0 -
Should keep in kind that uncorrected retransmitted frames are likely to happen inside of roughly a 3MB window (around 25mbits). At that level of service, you may not notIce it happeneing unless it is EXCESSIVE.
And with the AQM and other aspects of the 3.1 spec, errors can get 'masked' even further (in comparison to the impact on 3.0 connectoons).
That would be the greater concern... whether you are losing the benefits of OFDM and falling fully back to 3.0 protocols where errors are far more impactful.
0
