I have BeagleBone router and 3 EERO routers with Tread running on all of them. All devices have the same dataset. All devices are connected to the same Ethernet switch.
My EndDevices are connected to BeagleBone and EERO routers. From time to time I'm loosing connection from BeagleBone to End Devices. When there is red record (seen below) in routing table and I ping end device from BeagleBone bash with following command ping
fd27:fb0c:165:18df:e225:5320:b718:24e6 I'm not getting response but at exactly the same time if I will ping -I wpan0 fd27:fb0c:165:18df:e225:5320:b718:24e6 than I'm getting response.
When I remove red record from routing table with command ip -6 route del
fd27:fb0c:165:18df::/64 dev eth0
then both ping fd27:fb0c:165:18df:e225:5320:b718:24e6 and ping -I wpan0 fd27:fb0c:165:18df:e225:5320:b718:24e6 giving nice response. I found out that fe80::58ce:25df:5c8d:6341 is belonging to ETH interface of EERO wifi router which also has Thread running.
Any idea how that red record is added dynamically to routing table? Do you have any idea how I can avoid this situation? I would like to monitor devices with cacti which is installed on BeagleBone so I need to have stable communication link to end devices, but unfortunately once red record appear in routing table I'm loosing communication.
I will be appreciated for help with that.