Timeout while contacting DNS servers

[AdminOfThis]

I have a problem where the yahoofinance integration fails to fetch data one or two times a day, and therefore marks all tickers as 0.0$.

This is the error-log I get from HA:

Logger: custom_components.yahoofinance
Source: helpers/update_coordinator.py:424
integration: Yahoo Finance (documentation, issues)
First occurred: September 14, 2025 at 12:37:24 PM (2 occurrences)
Last logged: September 14, 2025 at 11:37:47 PM
Error fetching YahooSymbolUpdateCoordinator data: Cannot connect to host query1.finance.yahoo.com:443 ssl:default [Timeout while contacting DNS servers]

I understand that the root cause is most likely a local DNS issue, altough all other integrations and connections work fine.

Is it possible to just not update the sensors when a timeout occurs instead of "pulling them down" to 0?
Alternatively, add a "retry" argument to the configuration, so the query gets repeated before timing out?

[iprak]

Doe the requests recover?

The integration should do a retry after 20 second upon failure. There are other requests made to obtain cookie and it seems that those passed. Requests already have a timeout of 10 second so I don't want to add an intermediate re-try and complicate the call sequence.

The url noted in the failure is not the first request made by the integration so it is possible that this url is being routed elsewhere. In past, changes at Yahoo server end points have causes requests to fails in certain geographical regions.

Sensor values should not be updated if dataCoordinator runs into an exception, it would remain unchanged. Enabling debug logging from Developer tools for the integration might reveal more.

[iprak]

I was able to confirm that the entity value does not drop to 0. I was messing with network layout a few days ago and today when I looked back, there was no data for that period, it wasn't 0.

[txespirito]

In my case, I see these errors in the log.