#Suddenly having lots of DNS errors / cloud backups also not connecting

1 messages · Page 1 of 1 (latest)

patent heath
#

Just overnight started having failed cloud backups. DNS logs in home assistant appear to show a bunch of "refused connections" and repeated failed DNS lookups -- but the attempted connections are to the local docker container and so on.

Tried restarting. Here are some of the logs.

2026-01-28 07:54:56.213 WARNING (MainThread) [supervisor.plugins.base] Watchdog found dns plugin failed, restarting...
2026-01-28 07:54:56.215 INFO (SyncWorker_0) [supervisor.docker.manager] Cleaning hassio_dns application
2026-01-28 07:54:56.233 WARNING (MainThread) [supervisor.plugins.dns] Ignoring user DNS settings because of loop
2026-01-28 07:54:56.243 INFO (MainThread) [supervisor.plugins.dns] Starting CoreDNS plugin
2026-01-28 07:54:56.528 INFO (MainThread) [supervisor.docker.dns] Starting DNS ghcr.io/home-assistant/amd64-hassio-dns with version 2025.08.0 - 172.30.32.3
2026-01-28 07:55:01.721 ERROR (MainThread) [supervisor.plugins.dns] Detected a DNS loop in local Network!
2026-01-28 07:55:01.722 ERROR (MainThread) [asyncio] Task exception was never retrieved
future: <Task finished name='Task-598971' coro=<PluginDns.watchdog_container() done, defined at /usr/src/supervisor/supervisor/plugins/dns.py:348> exception=CoreDNSJobError('Rate limit exceeded, more than 10 calls in 0:30:00')>
Traceback (most recent call last):
File "/usr/src/supervisor/supervisor/plugins/dns.py", line 353, in watchdog_container
return await super().watchdog_container(event)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

also no longer have connection to the cloud:

any thoughts?

patent heath
#

so... not sure how home assistant got into this state - regular reboot didn't fix this. However I did a hard reboot of the system and appears to have cleared this up.

wide swift
patent heath
#

@wide swift regular DNS setup, only custom is I am also running tailscale and have the tailscale added as an additional 100.100.100.100 DNS server. A full reboot of the VM though seems to have cleared the DNS loop situation. partial reboot from inside developer tools ended up with many components failing to load.