NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.

Forum Discussion

ceh123's avatar
ceh123
Aspirant
Oct 05, 2023
Solved

New RAX70 router will not connect to unmanaged switch--seems to kill ISP Internet

Just installed a new RAX70 router connected to Spectrum ISP for Internet. Firmware updated.  When I try to connect an unmanaged switch (TP Link or Netgear) to the any ETH-out port on the router, it w...
  • ceh123's avatar
    ceh123
    Oct 06, 2023

    Actiontec MOCA AP is connected to Netgear router ETH port. 

    Discovered that root cause of the problem may have been a bad connector on the backbone ETH network infrastructure for the building. All switches were affected by the resulting intermittent connectivity, which just happened to coincide with the testing. Other parts of network were not affected by the backbone issue. Replaced connectors at both ends of backbone line. So far, no further problems with switches. Embarrassing example of the usual advice to check all cables. Did so on the local devices and also tested the backbone but missed the issue because of the intermittent nature and its coincidence with the switch testing. Not absolutely sure that this explains all switch "fails" but hoping it does. Thanks for all the replies, which definitely helped narrow the issue.

FURRYe38's avatar
FURRYe38
Guru - Experienced User
Oct 06, 2023

Is this Actiontec ether over Coax connected to the RAX router? 

 

I'd try a factory reset on the RAX and setup from scratch. Disconnect all but one wired PC and have a LAN switch at the ready to test with it connected to the RAX. After the router is setup, test with the wired PC with the router first. I presume this will work. Then connect the switch to the router and move the ethernet connected PC to the Switch and test. Problem still happen here? 

ceh123's avatar
ceh123
Aspirant
Oct 06, 2023

Actiontec MOCA AP is connected to Netgear router ETH port. 

Discovered that root cause of the problem may have been a bad connector on the backbone ETH network infrastructure for the building. All switches were affected by the resulting intermittent connectivity, which just happened to coincide with the testing. Other parts of network were not affected by the backbone issue. Replaced connectors at both ends of backbone line. So far, no further problems with switches. Embarrassing example of the usual advice to check all cables. Did so on the local devices and also tested the backbone but missed the issue because of the intermittent nature and its coincidence with the switch testing. Not absolutely sure that this explains all switch "fails" but hoping it does. Thanks for all the replies, which definitely helped narrow the issue.

  • ceh123's avatar
    ceh123
    Aspirant
    Oct 09, 2023

    One follow up discovery. We had a third factor. The Actiontec MOCA Ethernet-over-coax AP was connected to the router via coax, and delivered Wi-Fi and two LAN-out ports for hardwired ETH. As a MOCA device, the Actiontec AP had no ETH WAN liner in since the conversion from the coax to ETH took place inside the AP rather than via an external adapter. This AP was located near a wall-mounted ETH outlet, which connected to the router via an alternative Cat 6 ETH path. This outlet was not in use.

    While we were in the process of replacing the router and debugging the connectivity issues explained above, a new tenant moving into this space connected the ETH patch cord (left there by the prior tenant) from the AP LAN-out  to the wall-mounted ETH port (why?--perhaps to see if it would do something?). This created a network loop that crashed the network. Later on, as part of the testing sequences the line to this ETH port was disconnected from the switch that led to the router, which eliminated the loop. As testing continued, this line was plugged in and out, triggering multiple crashes. Then it was left unplugged until today, when it was plugged back into the switch--and re-created the crash. After unplugging this cable and rebooting the router, purely on a hunch we plugged the patch cord into the wall port and the LAN-out port on the AP and the network crashed. Bingo.

    Several tips here. We did have a loop problem but failed to see it until today. We facilitated the error by leaving the patch cord in the AP. The wall ETH port was intended to be used as a backup route to the office, but we left the port active even though we were using the MOCA route as the primary (for reasons not relevant here). As the office was empty, we never thought about the possibility that someone could have altered our test environment and inadvertently created a loop. 

    Glad we figured out this (hopefully last) piece of the puzzle!