Welcome to the Cox Community
Our community is a collaborative place to learn about Cox products and services. Jump into the conversation today!
Messages
Selected Messages
New Contributor
•
May 14th degradation in service
Hi,
I tried sending this through an email (took a guess at a couple support email addresses) but that didn't seem to work. Will post it here to see what folks think. I've put some images here: http://imgur.com/a/fmmUb showing the issue.
I've been a Cox customer in phoenix for 10+ years and have been using your internet service at my current home for the last 5+ years (new home build, I'm the first owner).
Over the past few months I've been experiencing intermittent issues with my service. Over the course of March and April (I believe) I had several service calls where everything was checked out, including replacing my cable modem and performing all of the steps associated with a "new install" (according to one of the service folks).
The symptoms of my issues consistently showed up only at night, and usually starting around 9pm. For a week or so, everytime it happened I would call in and report it. This was somewhat of a pain, though, since I only have a cell phone and because AT&T coverage in my house isn't great, I use an MCell to route my cell phone over the internet. In order to call in I have to disconnect that, and more than a few times my call dropped while speaking to one of your reps.
The performance I saw was at night my latency would skyrocket, I would start dropping packets, and even logging into Cox's speed test page would fail and the test would also fail. Usually after around 20-30 minutes the service would return to 'normal', but sometimes it took longer than that.
The performance got so annoying that I started tracking various points of data, gathered from my Cisco cablemodem's status page, along with pinging a few different hosts, including my router's default gateway. Anytime I started noticing problems with my connection, my ping times and latency data tracked it accordingly.
My final service call (I believe sometime in April) the technician told me that a "Service Issue" (or something like that) was going to be entered in so that another team, possibly people on at night, could look into my issue. After that happened, I called in a few days later to report a service problem again and the rep told me that the Service Issue had been resolved... it hadn't, since I was calling to report a problem again, but that was the point where my cell phone disconnected and I didn't bother calling back.
Fast-forward to May... on May 1st, around 7pm Arizona time, something changed. The consistent packet loss that I had been seeing had gone away and I didn't have any other issues. I'm not sure what happened, but it was a nice surprise and I truly felt like the problem had been solved.
Unfortunately, that lasted only until Saturday, May 14th at around 11am or so Arizona time. Since then I have been experiencing more consistent packet loss and latency yet again. It didn't seem to have gotten worse at night, but I've only had one night to experience it.
I've attached some snapshots of the data I've collected... I'm hoping this email will be helpful and actually get to the right folks. I'm not looking forward to restarting calling in whenever I have a problem... it's painful to walk each new rep through the problem, and just as painful to then try to find a time where a service call can be made, take that time off work, etc... I think it's pretty clear that there's something going on in the infrastructure, and nothing I need to be available for.
I'm not sure if a dialog can be done through email to track down the cause, but I'm more than happy to continue collecting the data (or expanding it), and reporting on what I've found. If I need to schedule a service call then I will.
Thank you,
Jason
Tecknowhelp
Valued Contributor II
•
What is the default gateway you are testing? Can you post a tracert to google(or other) while the problem is happening and when it's not?
Also, can you post your signal levels? Do you notice if your signal levels change when the problem happens? What was the other diagnostic data you got from the cisco modem?
Last, did anything of note happen May 14th? Trucks in area? Construction? Unusually hot/cold?
0
0
jh_phx1
New Contributor
•
Hello!
Thanks for the reply... I've added images of my downstream power and SnR here: http://imgur.com/a/kroGi Nothing of note changes on the 5/1 date where it was "fixed", and the big change in the graph occurred on 5/10 which is before the 5/14 "break" date. I also have the upstream signal and channel availability -- it's pretty much the same across the timeframe.
I do recall seeing a Cox truck in the general area working outside of the neighborhood, but don't recall the day exactly.
Nothing strange that I can recall on May 14th... weather was normal for phoenix.
The default gateway I'm pinging is the one reported to my router (70.190.45.1). (In case folks are wondering, I'm using a Ubiquiti EdgeRouter Lite, LAN to my computer via structured wiring)
For whatever reason it seems like the problem flares up around 9pm AZ time. I'll try to snag a traceroute tonight if it happens. Right now I'm seeing roughly 1 out of 10 ping packets getting dropped every couple minutes (script runs every 30 seconds, pings default gateway 10 times). Between May 1 and May 14th I'd go hours, sometimes days before the script reported any dropped packets...
Jason
0
0
Tecknowhelp
Valued Contributor II
•
May I ask how you collected that data? Almost looks like DocsDiag, but I didn't think that worked since Cox disabled SNMP on local. Edge utility?
Was the truck 5/10 or closer to 5/14? Wondering if they were trying to fix or causing the problem. Also, do you have time specifics on the 5/10 event? Also, looks like your signal first started to drop 5/1.
What is the coaxial situation from the tap(street) to modem? I know nothing changed there, but any possible points of failure?
Last, do you have any other Cox services? Have they been effected at the same time?
0
0
jh_phx1
New Contributor
•
I scrape my Cisco modem's status page (Cisco DPQ3212) every minute with a script I wrote and save all the values.
Don't remember the truck timing, unfortunately...
As for times... yes, specific times I see "fix" on 5/1 at 7pm AZ -- that's where virtually all my packet loss goes away. There's no noticeable change in other values at that time, but on 5/1 at 4pm (3 hours earlier) my ping times spiked for 30 minutes so theoretically that could have been when something changed.
"Break" occurred on 5/14 at around 11:15am... ping times started to spike for around 30 minutes again, and all of a sudden my packet loss is back. Signal, power, etc... show no significant change at that time.
Something happened on 5/10 to change how my downstream power looks in the graph... that is at 3pm AZ (again, ping spikes and packet loss but only for 10 minutes). But that change was during my "good" period, and it was another 4 days until the "break" occurred.
The tap (? -- the box for the neighborhood) is on my property, roughly 20 feet from where it goes to the box to my house and the straight to a wiring panel in my closet. I have the modem in that closet. I don't recall any work being done there during that time, but I'm not around all day so could easily have missed it.
Dunno... it's a puzzler. I need to add the ability to track my trace route and note the default gateway with each test I'm collecting, just in case any of that changes, but honestly what I'm hoping is that the date and times help Cox correlate some other maintenance or network changes. I don't know how many of my neighbors use Cox... i only notice the latency because I do gaming at night... I can see the lag and my voice comm becomes problematic (so I'm told).
Thanks again for the questions, I appreciate the help!
Jason
0
0
lvempknight
New Contributor
•
0
0
jh_phx1
New Contributor
•
0
0
Tecknowhelp
Valued Contributor II
•
Darn, thought you found some way to get SNMP to work. I have to stop being lazy and make a script of my own. I may wait for my 16x4 TM1602 eMTA though.
My guess is they added a amp for some new construction and it took them a while to balance the levels. Hopefully its that and not intermid noise that might come back. ::fingers crossed::
0
0
jh_phx1
New Contributor
•
Ah interesting... that could be, there is a bunch of new construction around the area... not quite next door, but within a reasonable distance for sure (new neighborhoods).
So far it's been staying good... I'll be crossing my fingers, as well.
Out of curiosity... I'm totally not familiar with the modem side of things, so curious about your TM1602 comment... how much does the modem matter these days for connectivity, throughput, etc...? I have the Cisco one via Cox and never thought about swapping it out for any reason other than it gets fried :) Are there reasons I'd want to think about something different? I.e., why are you waiting on a new device?
Thanks!
Jason
0
0
Tecknowhelp
Valued Contributor II
•
One manufacture vs another doesn't matter that much, but the number of downstream channels it can bond on can. Its a good idea to have the modem have a maximum data rate double your max provisioned speed. Premier goes up to 125Mb x 2 = 250 > 343Mb with 8 channels. So don't worry about it unless you have Ultimate.
0
0