[AR] I am within Parameters but massive uncorrectables, WHY?
Based on the information provided »Cox HSI Forum FAQ »What should my modem signal levels be? I am within the parameters for Cox HSI; so why are my uncorrectables in the the hundreds of thousands in 2 days?
SB6180 Connect Screen Approx 10:30
SB6180 Connect Screen Approx 11:30
My COX HSI Service is:
Premier 28mbps/5mbps (+7mbps SPEEDBOOST) and Premier Phone Service.
HSI Motorola SB6180 and Phone ArrisTM502, Router is Mikrotik RB751U-2HnD ROS v5.15
(phone works every time I call for help, just fyi)
When at my 4656 address, the Cox Service was stellar using doc sys 1 SB modem, and Mikrotik ROS v4.17 and Arris Phone modem. I had phone, internet and TV. It was wonderful for a year straight. Called Cox for changes to service and once after a storm. We moved at the end of our lease to 111 address. The nightmare began that March and continued into the late fall. We had to change to Directv do to pixelation and loss of picture. Over the winter months, everything worked fine after buying several different routers. Linksys(Cisco), DLink, Mikrotik and being given a Motorola SB6120. In February the weather was unseasonably warm and KAPOW, all the issues (issues will be listed below) started all over.
After about 3+ truck rolls and visits we were able to function providing we did the following:
Downstream - Channels (4) If any of the dBmV's went above 3 or -3 then all new internet requests would fail. Skype voice would be choppy and video would DC. Webpages would at best partially load and mostly text only.
UpStream - Channel (1) If it reached 50 dBmV then Webpages would time out and server not responding and uploads to college platform would fail. Skype would not function and gaming would be extremely laggy and ventrilo would hit 7000ms or higher latency.
The fix to this was a simple two way splitter than could move the signal levels back to a 0 dBmV on the downstream and to 47dBmV on the Upstream. This would need to be done only once a week at first then moved to 2-3 times a day.
On the Monday last week the 21st of May, if you can believe it everything got worse, much, much worse. Trucks rolled in, 3 to be exact. Please understand that my calls to Cox Technical support required these trucks to roll regardless of the information I was trying to provide them on the call. I have offered to all the Support Personnel when I call, including "supervisors" the log screen shots of my modem logs so they can see the errors.
These errors include some of the following:
SYNC Timing Synch failure - Loss of sync
Lost MDD Timeout
Cable Modem Reboot due to MDD Loss
TLV-11 Unrecognized OID (this one is after any power up of modem)
No Ranging Response received - T3 time-out
TFTP File Complete - but missing mandatory TLV
ToD request sent - No Response received
TFTP Request Retries exceeded, CM unable to register (this one always requires us to reboot the modem and pray to a higher power)
I am not genius, but I am fairly certain that with the Main line showing an inverted tilt (possibly reverse tilt), and with problems browsing when in the "acceptable signal level range" that there must be a bigger problem.
My experience with DOCSiS1 has most Cable Services with similar signal parameters of:
Downstream @ +7 to -7 dBmV
with SNR @ 36+
Upstream @ 43-51 dBmV
My current levels as of this initial posting are:
Acquire Downstream Channel (807000000Hz) Locked
Downstream @ (807000000 Hz) 2.0dBmV and SNR39.6
Upstream @ (36000000 Hz) 44 dBmV
-Downstream by Channel-
Chan 1 (807000000 Hz) 1.9 dBmV SNR 39.3
- correctables 458270, uncorrectables 92364
Chan 2 (813000000 Hz) 2.0 dBmV and SNR 39.9
- correctables 5, uncorrectables 0
Chan 3 (819000000 Hz) 1.9 dBmV and SNR 40.0
- correctables 11, uncorrectables 2
Chan 4 (825000000 Hz) 1.6 dBmV and SNR 39.8
- correctables 21, uncorrectables 2
Chan 2 (831000000 Hz) 1.3 dBmV and SNR 39.9
- correctables 10, uncorrectables 7
-Upstream by Channel-
Chan 1 (36000000Hz) 2560 Ksym/sec 44.0 dBmV
Speed tests prior to the 21st of may done on the cox.net site Kansas server.
Pings 37 to 66 ms
34-66Mbps/6-8Mbps depending on the signal levels. If they were at 0 DS and 46 US then 100+Mbps spike averaging to the 60's happened, if at the +/-3 or 50 then if it worked in the 30/6.
After the 21st (You are configured correctly and it is working as intended), I can no longer fix any issues myself by adding or removing the splitter. the following are the Cox.net site speed test results.
12-41 Mbps/7-8Mbps if you can get it to work. Times out with a download only of 2-3 Mbps / no data and ping 90 but doesn't save the data in the log. Actually, only about 2 in 5 tests show up on the log. Have screen shots of all this if anyone wants them.
I am certain the uncorrectable being 1/4 of the correctable; and the correctable being in the hundreds of thousands is definitely not in the acceptable parameter range.
So what is my next step?
I power cycled the modem and router properly 5-27 @ 10:15AM and in just 1 hour and 30 minutes I am at almost 10000 correctable and 2300 uncorrectable on channel 1. The screenshots are at the top of this post.
When you look in your modem's log do you see a lot of T3 errors?
First image is from the SB6120 (Replaced on 5-21 @ 5pm)
Just an example
SB6180 - 5-27 reboot
Second image is from the SB6180 (Rebooted on 5-27 @10:15am)
Yes there are a lot of t3 Time Outs. I am convinced that somewhere on the Main Line (not my home, not the drop line from the Pedestal to my home but the line into the neighborhood or the pedestal itself) has an issue with either a bad connection, or faulty hardware.
ok - you mentioned phone service in your original post. Do you still have that with the cable co? and if so, is it working correctly when you're experiencing problems with your data?
Irish SharkPlay Like A Champion TodayPremium,MVM
Las Vegas, NV
reply to WiFiTech
You have an upstream issue in that your MODEM can not "talk" loud enough for the CMTS to hear it.
Call Cox and have them run some tests.
"You can observe a lot by watching". Yogi Berra
reply to bdnhsv
Image is just an update on the correctable/uncorrectable @ 12:45pm
SB6180 12-45pm update
Outside the home is a splitter (unsure what is on it now, used to be a 4 way with only two lines out). One cable (RG6 installed around 3-12-12) goes to the SB6180 Modem, and one cable (existing from to moving in, 3-1-2011, home ~6 years or less old) going to the Arris Modem supplies Phone into homes existing wiring.
Yes I am able to talk to the Cox Technicians all the time that the service is having issue, modem is rebooting on its own and even have a clear conversation when they are telling me the phone modem is at 50% packet loss. I had to ask the CSR to verify he was on my account and looking at the ARRIS after telling me there was 50% packet loss on VoIP.
reply to Irish Shark
I love you man, but I have been on the phone with them at least 3 times since Friday 5-25-12. They just want to roll another truck to my home. A friend of mine that works for Cox is having his Supervisor (Has background working in the Rogers, AR area) look into the issue for me. If that doesn't work then I may have to head back to Wichita and see if any of my old FIEST friends work in Cox's Management. I am semi retired, and may just have to update my certs and find my FLUKE meter and bill them $400.00 an hour for consult to fix the issue. I have a roll of LMR 400 still from my days doing WiFi installs @ Marriott Courtyards.
reply to Irish Shark
The responses to my inquiry for a LINE TEST is that everything looks to be in the acceptable range. They are sorry that I can't stream video, audio or use skype to make a call and that I am getting disconnected from my MMO, and uploads to my colleges platform fail. However, those are not certified uses on their system. Thank you Eugene for the help, but the correctable/uncorrectable counts and the MDD reboots and the T3 timeouts say it is not a problem with my home network.
wifi - my first guess would be impulse noise, but the fact that your arris emta is functioning when your sb is not kind of throws me off a little. Have you spoken with any neighbors to see if they are having similar issues? Also might help if your friend could get the headend tech to fire up a spectrum analyzer and take a look at the node that provides your service.
My friend is 5 hours away in Kansas. Also, another point was brought to my attention with regards to the SB6180 and the connect interface being open continuously. The correctable/uncorrectable may be misrepresented. After the reboot this morning @ 10:15am everything seems to have evened out. I am going to do some of the usual stuff, upload to the college platform, use Skype, stream A/V and play my MMO. I will let you know tomorrow how everything goes. Unless it all starts to fail again. Thanks.
reply to WiFiTech
With all the information that has come in both posted and PM, I decided to try a few things. Around 2pm yesterday I stopped worrying about the connection details and just went back to doing what I do.
SWTOR - MMO latency below 100, and only lost connection 2 times
Skype - Didn't get the chance to use it.
Cox Phone - didn't get the chance to use it.
Browsing - Did a little, craigslist noting over graphical.
Pandora - Streamed music while I was MMO'ing without issues.
I checked on the correctable/uncorrectable after 8 hours they were far less than was witnessed with the SB6180 modem interface open and monitored.
@ 12am 46174/9769
8 hours ~274 / ~30
That is definitely better.
I went ahead and rebooted the modem from the interface @7:55am and will see how the connection performs.
DownStream Channels ~ 6.2 with SNR ~ 39.5
Upstream is ~42.5.
cox.net speedtest 69ms; ~66Mbps/8.2Mbps
In just the last 1 hour and 30 minutes correctable/uncorrectable ~350 / ~120
Much more realistic numbers in regards to what is happening and definitely more believable in regards to having anything working. There are several people offering assistance in PM's and I will post the results of what we try.
Well them codewords counts is not bad at all, and really nothing to worry about. If I was you I would focus on the modems log page, where the modem logs any errors that happen. Which if the modem is rebooting, or there is other problems. There will be errors in the log page, with a date, and time next to it.
WiFiTech looking at the above post of yours with the errors from your modem log. T3's, and T4's are a sign of upstream trouble when they are appearing often. Which are bad, and can cause connection issue's, including modem rebooting randomly. My suggestion to you would to probably continue monitoring things yourself, and if problems continue. Maybe send CoxTech1 a PM on here, he will need your modems mac address. But he can look at things from cox's end, and suggest what he thinks should be the next step in getting your issue resolved.
reply to WiFiTech
Thanks Lilstone87. I was on the phone with Cox, "Ben" this morning and we did a few things to get a better handle on it. It seems that the upstream power is increasing and there is a pattern. He is going to check into a few things. We also looked at the history and we are noticing that the Arris Modem is not fluctuating at all like the SB6180. I will update again after Ben gets back to me. He and I are in agreement, there really isn't anything in a that can increase the upstream by 2 to 3.
SB6180 - 5-30 reboot
Modem powered off two times today. ~ 30 minutes ago. (image above)
DS Frequency 819, SNR 39.4, Power Level 5.0. US 42.5
reply to WiFiTech
If you want to see if the problem is resolved you can follow the blog style thread @ »[AR] AR-Rogers-Post Meadows Dr. 5-31-12 to date..
reply to WiFiTech
I am ending my posting to this thread. I am going to start a new thread that I am going to use just to document everything going on so that I can just refer the Techs/Supervisors and everyone else that gets involved cause either I am the cause of it all or the service is just not fixable or misinformation is being spread all over.
If you want to see if the problem is resolved, you can find the blog style thread @ »[AR] AR-Rogers-Post Meadows Dr. 5-31-12 to date..