Jump to content

Fathomless

Members
  • Content count

    86
  • Joined

  • Last visited

  1. Fathomless

    XR500 and Ethernet Switch

    Good choice it should help you with Ethernet stability, since you would offload switching function to another device.
  2. Fraser, thank you for the help with Netgear fix. ...up time is 36 days with .18 firmware.
  3. I see, that is IP address you acquire from Comcast. Comcast would normally give you IP for 4 days. Does it break every 4 days and do you see ip address switch frequently? Expansion of lease duration should not cause interruption in service. XR500 works as client in this case with Comcast DHCP service, have you contacted Comcast for the service issues? it might be DHCP Server or might be a modem...
  4. The issue might be different in your case. So you loose connection only from one host at the time during lease change? (do you have other devices using the same IP for example static IP?) If you reserve IP address on the router or provide static address to the host would it mitigate connectivity issue?
  5. Fathomless

    Cm1000

    What about active queue management (AQM)? http://www.cablelabs.com/wp-content/uploads/2014/06/DOCSIS-AQM_May2014.pdf
  6. Still running debugging firmware, with good results: Uptime: 26 days, 21:23, load average: 4.36, 4.26, 4.21 let me know if you want me to test beta firmware.
  7. sounds good thank you. For now I will run with debugging firmware, it will be close to 25 days by June 3rd.
  8. I would like to try it new firmware, but most likely one week later. Can start testing on June 3rd.
  9. debugging firmware - Uptime 05:34:41 up 16 days, 11 min, load average: 4.39, 4.20, 4.12
  10. I think Fraser will be online in 8-9 hours, ask him about debugging firmware if you are willing to participate in debugging. Side effect in my case - i cannot repro that failure. . ....but if you just wanted quick mitigation - disable instant-on.
  11. The issue here is well isolated already netgear just does not want to repro it, or they try to repro with debugging firmware, which by default is stable and has networking stack altered for capture . So if I would try to repro it now I would say no repro... but if that is true what stops netgear from incorporating additional, software into factory firmware? Issue happens when Xbox is in instant-on mode (in hibernation), there will be frequent port reset on the switch triggered by xbox. Frequency is 1-2min. the pattern will look like this: 10/full(~1min) -> disconnect/reset(1-5sec)->1000/full(1-5second)->10/full(~1min)->disconnect. What instant-on-disable going to do? it removes the pattern specified above, so weakness of XR500 is not exposed anymore. If you see disconnect after instant-on-disabled that means issue is different. if I was a netgear engineer in the lab I would connect 4 Xboxes into 2 XR500, put Xboxes in hibernation and capture repro all day long. Or create generator of pattern and connect generators into XR500, increase frequency and make it 15 seconds per cycle, that could be done by xbox team or netgear team. Did you talk to tier 1-2 support or to premier support for xbox? Tier 1-2 would be vendors at Microsoft, they will not have incentive to bring that issue to the product team, and would be limited by SOP in troubleshooting.
  12. Fathomless

    Cm1000

    I do not use any attenuation, did not want to reduce downstream signal. The difference in signal is significant 3dbmv. also +1 on heat dispersion Sb8200 runs significantly hotter compare to CM1000, heat is a problem particular when dust starts building up in a year or two, it will work like insulator, overheating the components...potentially. I see why you can select sb8200 or mb8600 though all 3 modems have the same chip set. SB8200/mb8600/CM1000, the difference comes down in number of ports, memory and storage. I think by the time people will start to use more then 1port, there will be better technology new modems most likely, even if you do not use memory it will still consumes power, so more not always better.
  13. Fathomless

    Cm1000

    Not necessary it will depend on signal to nose ratio i am almost at 40db, and another one is that CM1000 has better out of the box stats, try to ask your provider to attenuate in line amplifiers, so your signal can go up. +1 towards CM1000 on the spec- specification calls for 38-48dbmv compare to 45dbmv-51dbmv(that could be achieved in the lab for many users who has Comcast, but note that Comcast has a range 35-55dbmv with automated adjustment of upstream if signal quality is low), you will be surprised that Comcast modems work with 33dbmv, and do not drop connection. I will give you a few stats and you can compare... my case: SB8200 signal to noise 39Db Downstream average: 0-1 dbmv upstream average: 36-38dbmv CM1000 signal to noise ratio: 39.5 Db downstream average: 0-1dbmv up average: 40.5-42dbvm in my case provider was not willing to regulate power on the modem or do any changes to the line amplifiers, and arriss was not willing to investigate disconnect issue, since upstream signal was bellow 45dbmv. I was not willing to put directional attenuation on the modem, since i have perfect downstream.... there is probability that sb8200 had ARP table corruption, since even 192.168.100.1 was not accessible issue is similar to Ethernet ports disconnect ....similar to XR500. So SB8200 has too many points of failure
  14. Fathomless

    Cm1000

    CM1000 is good choice, it provides good reliable connection, if your area supports docsis 3.1, go with it. Tested it with 150mbps and now upgraded to 1gbps.(when i had sb8200 and 1gbps it was disconnecting twice a day) CM1000 is superior to SB8200 and MB8600 why? - they fixed the bug with random disconnect, when arris sbb8200 has low upstream signal - bellow 36dbmv - it will disconnect randomly. - customer support for CM1000 exist, where with SB8200 does not, you will be running around and listening automated manual, Motorola does not work with customers very well, Motorola MB8600 is out due to customer support issue.(Arris operated by Motorola)
  15. Tested 1Gbps for 2 days, router is still stable, with 1gbps speed I still have a limit at 1Tbite a month, I used 10% of that in two days, while testing But I have positive results on the setup and choice of equipment. Just to remind you guys I downgraded to 150 in the past just to troubleshoot loss of internet, arris sb8200 was not stable. Tested xr500 via 5Ghz wifi. - at close proximity, 1 wall/floor away from XR500 I get speed tests results as 540-600Mps; where R6700 gets 300Mps max, I think that is ok I do not expect anything more, and I am ok with lowest denominator. - with 2 WIFI devices I use the same SID on 5Ghz, in all corners of the house speed is at least 290Mps; wifi coverage: house with garages approximately 4800Sft - did concurrent tests as well could not kill "debugging" firmware. I will wait for 15 day mark, and would think that debugging firmware fixed my issue with LAN ports frequent disconnect. the WAN port I am not concerned about since NG CM1000 has stable connection to the router, way better compare to Arris SB8200.
×