Network setup:
AT&T gateway set to ip passthrough mode going to a Turris Omnia (TurrisOS 5.3.2). Turris Omnia to a 16 port Netgear switch. I have a tp link router running as an access point with OpenWrt (19.07.7).
All non-guest devices have static ip addresses assigned. Guest devices are assigned an ip via DHCP.
In the last few weeks, my network seems to randomly drop. Sometimes it will come right back up and other times some of the devices are assigned 169.* .* .* ip addresses. I’ve noticed the devices that receive the bad ips are wired devices and wireless devices that are currently connected to the access point.
A few things that may be red herrings:
-
We mostly notice the drop in the evening, which is when we are watching our AppleTV 4k. Right after some of the drops this shows in the logs:
Dec 24 12:19:35 turris dnsmasq-dhcp[8274]: DHCPREQUEST(br-lan) 192.168.1.199 f0:b3:ec:XX:XX:XX
Dec 24 12:19:35 turris dnsmasq-dhcp[8274]: DHCPACK(br-lan) 192.168.1.199 f0:b3:ec:XX:XX:XX Apple-TV-Living-Room
Dec 24 18:19:35 turris /dhcp_host_domain_ng.py: DHCP update hostname [Apple-TV-Living-Room,192.168.1.199]
Dec 24 18:19:35 turris /dhcp_host_domain_ng.py: Refresh kresd leases
I have found a few other posts on the internet referencing AppleTV logging these same lines, but the posts specifically mention that the logging is the only issue - no drops happen.
I want to stress that this only happens during most of the drops, certainly not all of them. The Apple TV has an assigned ip for both wireless and wired. -
If the wired devices are not working after the drop and unable to get an IP address (169. * .* .*) correct ip addresses are assigned if I just restart the network switch.
-
I have IPv6 settings all disabled on both the router and access point, but some devices still show a ip6 address next to the ipv4 address on the overview page in the associated stations table. The Active DHCPv6 Leases table is empty. What could be assigning the ipv6 addresses?
My network has been stable for almost a year until the end of November or so. I can’t think of anything that has significantly changed over that time besides additional devices being added. Is it possible the switch is bad and causing the drops? Is something in DHCP all of the sudden causing issues? Any ideas or places to start would be great.