HA unable to check for update in UI

[Skye]

My HA instance:

Is unable to check for updates for core, e.g. I was running 2025.8.0 and it did not show me there was an update to .1

This started happening when I changed ISPs and router hw. I had to change the DNS server in the config page to be able to force update via the terminal as the Update page in the UI did not give any error or message when I hit the check for update button, but the terminal cmd line did say that the update endpoint was unreachable (very weird but OK).

Currently the only way for me to update HA is on the cmd line and EXPLICITLY passing on the version number, if I just use “ha core update” it doesn’t see the update either.

I do not get any error messages in the syslog that I can see.

If anyone is interested in debugging this let me know what I can do as this most certainly is not the happy flow

EDIT:

I do see some error logs in supervisor and host:

2025-08-12 16:12:44.243 homeassistant dockerd[462]: time="2025-08-12T16:12:44.243061225Z" level=error msg="[resolver] failed to query external DNS server" client-addr="udp:192.168.0.108:45283" dns-server="udp:195.130.130.2:53" error="read udp 192.168.0.108:45283->195.130.130.2:53: i/o timeout" question=";checkonline.home-assistant.io.\tIN\t AAAA"

Strange thing is that the DNS server I configged for IPv4 in the network page is 212.71.0.33 which is not the one HA seems to be using looking at the log

[NathanCu (Nathan Curtis)]

Read the supervisor log if you see ‘supervisor no internet’ that is what’s causing the failure of update detection. Troubleshoot THAT. (you almost certainly have either a time sync or DNS issue if you saw this in conjunction with an ISP change happening)

[Skye]

2025-08-12 18:12:36.242 WARNING (MainThread) [supervisor.jobs] 'GitRepo.pull' blocked from execution, no supervisor internet connection
2025-08-12 18:12:36.242 WARNING (MainThread) [supervisor.jobs] 'GitRepo.pull' blocked from execution, no supervisor internet connection
2025-08-12 18:12:36.243 WARNING (MainThread) [supervisor.jobs] 'GitRepo.pull' blocked from execution, no supervisor internet connection
2025-08-12 18:12:36.243 WARNING (MainThread) [supervisor.jobs] 'GitRepo.pull' blocked from execution, no supervisor internet connection
2025-08-12 18:12:42.243 WARNING (MainThread) [supervisor.jobs] 'GitRepo.pull' blocked from execution, no supervisor internet connection
2025-08-12 18:12:42.244 ERROR (MainThread) [supervisor.store] Could not reload repository 982ee2c4 due to StoreJobError("'GitRepo.pull' blocked from execution, no supervisor internet connection")
2025-08-12 18:12:42.244 ERROR (MainThread) [supervisor.store] Could not reload repository 5c53de3b due to StoreJobError("'GitRepo.pull' blocked from execution, no supervisor internet connection")
2025-08-12 18:12:42.244 ERROR (MainThread) [supervisor.store] Could not reload repository core due to StoreJobError("'GitRepo.pull' blocked from execution, no supervisor internet connection")
2025-08-12 18:12:42.244 ERROR (MainThread) [supervisor.store] Could not reload repository a0d7b954 due to StoreJobError("'GitRepo.pull' blocked from execution, no supervisor internet connection")
2025-08-12 18:12:42.244 ERROR (MainThread) [supervisor.store] Could not reload repository d5369777 due to StoreJobError("'GitRepo.pull' blocked from execution, no supervisor internet connection")

Looks like it. As updated in my original post, I have no clue where HA is getting that dns-server=“udp:195.130.130.2:53” from though

[NathanCu (Nathan Curtis)]

How is your network configuration setup?

[Skye]

No IPv6

DNS is the official DNS ip provided by the ISP and pingable in the HA terminal

[NathanCu (Nathan Curtis)]

Ok then that means you don’t have your own internal DNS there… See what adding 1.1.1.1 as first DNS (you can drop your or your ISP as secondary or tertiary or leave it out entirely )

[Skye]

Adding the Cloudflare DNS as the primary solved the issue, thanks!

[WallyR (Wally)]

Have you restarted the HA host (not just HA core).
Many of the network stuff run in the OS and core just relay to the OS, so if the OS is at fault, then restart core will have no effect.

[Adonis_VII]

I also had this issue, and these resolved it (adding Cloudflare and restarting hardware not just Core)