Forum Discussion

enrapt's avatar
enrapt
New Contributor III
5 years ago

Ping & Packet loss [Megathread]

Collecting the threads where people are saying they have ping and packet loss issues.

Unlikely that this wide spread problem is caused by a "router", "signal issue" or "bent cables" (unless somehow everyone's cables got bent over night across the country? Santa Claus?) These problems are ranging from California to New Orleans, Arizona and Nevada between. 

I know these are unusual times but also sending technicians to check cables or trying to blame modems, house cabling for the issues doesn't help.

What can you do:

  • Install PingPlotter (14day trial) and start collecting data. 1.1.1.1 and 8.8.8.8 are Google's anycast DNS servers that should be optimized for latency. You could also try to ping cox.net, zoom.us, google.com or try to find your game server addresses. Try to collect several days of data to see how the service degrades in certain peak times
  • Post the results here and email cox.help@cox.com with the data and your issues/location etc. See if you get an answer or anything helpful to address. In some cases they want to send technician over but if you already had technician check your signals or cables, it's likely not going to help
  • If the problem persist, consider filing informal FCC complaint as azchips did on his case

Cox, what can you do:

  1. Escalate these issues to your manager/director customer support who can then escalate it further, hopefully up to the CEO/whoever is the director of Cox Communication. 
  2. Acknowledge that the issues exists and communicate to customers what is happening, what you plan to do and when. Email customers affected or post in this forums or on twitter.
  3. Launch your own investigation to the network look for bottlenecks, speed up your upgrade plans if you can. Even after this crisis, we are likely to see increase in businesses and consumers internet usage. 
  4. Focus improving the latency in the overall network. This is what most people are complaining about it. Since affects how fast new connections open, dns requests get resolved, and providing good experience for applications that require stable and constant connectivity (video calls, ssh, gaming). 

Threads (most of these are from the last couple of weeks):

  • enrapt's avatar
    enrapt
    New Contributor III

    Update Apr 1

    • Added few more threads on the list
    • Chatted with customer support over twitter. Sent more traceroutes. Part of the time they were confused about the hops and why my second hop is 10.134.0.1 and is that part of my network. (It's not, it's part of theirs and lot traceroutes from different customers here show 10.x.x.x.x address as second hop). Didn't make any progress and didn't receive any new information. They just asked me to reset the Cox modem to factory settings and offered to send technician to look again. (Technician visited 3 days ago and found nothing wrong expect their network likely overloaded).
    • I monitored Zoom and the latency to AWS (which is what they use for call). When I speak I often get "Unstable connection" and the packet loss is up to 25% on upstream (sent).

    I think their issue is that the cable (not fiber) is not optimized for lot of upstream usage. Now with people doing more video calls etc, their upstream is likely more congested and can't handle all the traffic increasing the pings and packet loss. Not sure what kind of upgrades they could do to increase the upstream capacity. 

    • enrapt's avatar
      enrapt
      New Contributor III

      From Cisco's manual (Troubleshooting Slow Performance in Cable Modem Networks): https://www.cisco.com/c/en/us/support/docs/broadband-cable/cable-modems/12551-troubleshooting-slow-perf.html#upstreamchannel

      Upstream Channel Congestion

      The upstream channel is normally the most precious resource in a cable system. Currently, most cable service providers use a 1.6 MHz channel width and Quadrature Phase Shift Keying (QPSK) modulation in the upstream path. This equates to approximately 2.5 Mbps in total available upstream bandwidth for all users connected to the one upstream channel. It is important to ensure that the upstream channel does not become over used or congested, otherwise all users on that upstream segment suffer poor performance.

      If upstream channel usage is consistently above 75 percent during the peak usage time, end users begin to suffer issues such as latency, slower "ping" times, and a generally slower Internet experience. If upstream channel usage is constantly above 90 percent during the peak usage time, end users experience an extremely poor level of service because a large portion of the end user's upstream data will have to be delayed or discarded.

      Sounds like we are in the 75-100% usage.

  • enrapt, thank you for this. Ill also be posting additional information once I can collect it showing I escalated to cox executive escalation branch this infrastructure issue over two years ago and nothing was done. Hopefully people might be able to use that in cases to drive a resolution from Cox. 

  • Codey's avatar
    Codey
    New Contributor

    I'm experiencing these exact same issues now. Started noticing it on 03/31/20 while playing Hunt: Showdown. I can see consistent lag spikes from 50-450ms on my windows desktop (wired), Macbook Pro (Wireless), and my brothers PC (Windows 10, wired).

    We both experience rubbe rbanding, lag, and lost packets on a wired connection. Have  tech coming out tomorrow, so we'll see what he find. Can't wait to be charged when the guy can't find any issues and blames it on my equipment (Oh wait, I already am. I had to play for the "Customer Care" Program for the next 90 days, so I guess its better than the inevitable $75.) 

    I swear, it would be amazing if Cox actually had competition in my area; maybe then they would be incentivised to improve their services, or offer better prices. I'm currently getting raked over the coals with fixed high prices, while my friends right down the road benefit from unlimited data, 1Gbs right at, or less than I pay for a fraction of that. 

    Cost complaints aside, I'm having ping issues too. 

  • sphayes01's avatar
    sphayes01
    New Contributor

    Thank you for this, I am having the same issue but I have not made a thread. High packet loss shown in PingPlotter. Hopefully this thread will cause some action.

    From what I've seen, Cox is saying the infrastructure can handle the traffic just fine, but I disagree.

    • AZchips's avatar
      AZchips
      New Contributor III

      Add me to another frustrated Cox HSI customer!  After dealing with the packet loss and high jitter issue for weeks, I went ahead filed FCC informal compliant on last Sat (7-day PingPlotter tracegraph attached showing problem starting from Hop2).

      https://forums.cox.com/forum_home/internet_forum/f/internet-forum/25181/az-chandler---constant-packet-loss

      Today I received a call from Cox rep in their Exec Office.  He admitted that the node over-saturation is indeed the problem.  He said my neighborhood node has 500+ people on it while typically should only have 200-300 (I'd argue it should be even lesser than that).  

      I asked him when the upgrade (a note was left from Cox in my Cox Inbox indicating "infrastructure upgrade" was coming; this was Sep 2019) could happen, and he said it was supposed to be on 02/19/20 but for some reason it did't.  He promised to get back to me on the schedule update.  We'll see. 

      What worries me - even after node split, there is still 260+ people on the same node...  I really don't think Cox infrastructure capacity is keeping up with the greater Phoenix area population growth.

  • paulsaz's avatar
    paulsaz
    New Contributor

    VOIP is not working well at all using and obi202. The packet loss makes the callers voice drop in and out. 

    • jonathonjoseph's avatar
      jonathonjoseph
      Contributor

      That is from whats called line jitter. Due to cox not having sufficient infrastructure and over selling bandwidth packages on infrastructure they knew could not support that many users tied into one CTMS/Node, your data packets get delayed in timing at the CTMS/node and even packets will be lost all together. Nothing you can do to fix it on your end, takes Cox doing infrastructure upgrades. Which at this point it appears has been neglected on a very large scale, they likely cant even resolve this in a timely manner if they wanted to. It will be interesting to see if anything comes of this as more of this is surfaced. I am almost willing to bet they neglected the upgrades only to build their fiber network to start charging more for fiber and push customers off cable and hope that mix balanced itself out without having to spend any money on the cable network. Which likely would of worked since it would of isolated packet loss and jitter between 5pm and 10pm giving them a viable excuse that is easier to explain it all. But now with shelter in place orders, and networks being used at the capacity in which they sold but did not actually have, well the curtain essentially has come down. 

  • dchadd's avatar
    dchadd
    New Contributor III

    Same here in northwest las vegas, FCC complaint filed, have a tech coming over tomorrow for a line verification. Before the covid crisis it was happening daily between 5PM-10PM. Now it's more like noon to midnight. odog on dslreports said there was some construction commencing on 3/11 on my leg of the cox network (i assume a node split). Still hasn't improved. odog said it was possibly was put off because of the covid pandemic and is checking for me. Hopefully they find something tomorrow, but doubt it. Century link fiber is across the street from me. I'd love to switch.

    Here's my original dslreports thread:

    https://www.dslreports.com/forum/r32648472-NV-Another-Night-of-Cox-In-Northwest-Las-Vegas?group=NV

    • jonathonjoseph's avatar
      jonathonjoseph
      Contributor

      I would not account for it being put off on the COVID crisis. Cox used all resources in Arizona to install fiber to offer new packages versus maintaining the current network and then continued to sell packages they cant support on nodes heavily over saturated. My nodes 85% over saturated and that came directly from an executive escalation case manager. Yet I can call in right now with my neighbors address who does not have Cox and ask to sign up for gigablast and they will sell it to me no questions asked. 

      • dchadd's avatar
        dchadd
        New Contributor III

        They should stop selling gigablast immediately, downgrade everyone back to 300 and admit to making a mistake.