-
Notifications
You must be signed in to change notification settings - Fork 11
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Local query not being triggered #33
Comments
Also this report seems to show I am not alone with this issue, since the start of 2023. I should also note that although I had configured for local fulfilment for some time before this, but given it seemed to just work, I never looked into that much until I recently was investigating further. I presume it was previously working, but it could have just been falling back to cloud operation, so I can’t be sure. Local fulfilment seems very finicky, which is a shame, seems like such an opportunity! |
I can't really speak for NabuCasa. But mikejac/node-red-contrib-google-smarthome and andrei-tatar/node-red-contrib-smartnora are having the same problem. This seems to be a problem on Google's side. A fix is expected mid-May. See https://issuetracker.google.com/issues/261893570#comment31 |
Thanks @Caprico85, that looks like the issue, I appreciate the heads up. I may investigate how to block my Google devices from upgrading their firmware in the future once I get software stable, given Google’s poor responsiveness. |
Google released a new firmware for their speakers, which fixed local execution. At least for node-red-contrib-google-smarthome. I guess it's the same for NabuCasa. |
Environment:
As above, I have configured each of the components as per the instructions and when debugging the google side, I can see successful
IDENTIFY
andREACHABLE_DEVICES
intents but theQUERY
andEXECUTE
handlers are never invoked, despite trying multiple permutations of the following:SYNC
s)PROXY_SELECTED
, additional logs, reverting to simple examples)The Google devices seem to have flagged each reachable device as without a local path and always fall back to cloud, without even trying. This is despite the fact they happily connect to the HAAS instance via the MDNS announcements, which are fully resolvable from devices on the same network. My nginx reverse proxy sees the fallback
EXECUTE
and multipleQUERY
requests from the Google cloud at each reachable device operation. No attempt is made to run these locally. If I disableSupport local query
in my google action, the fallbacks occur even when the Google device is placed intoLOCAL
mode. IfSupport local query
andLOCAL
mode are both enabled, the Google device reports that each reachable device is not available and I have to return the Google device toDEFAULT
mode to have the fallback work again.I note that others have had this issue with the Google firmware that I am on (1.56.324896) here and not when at least one device was on an older firmware, perhaps that is the issue? Can anyone confirm they have local fulfilment working with this (and only this) Google firmware?
So to me, it looks like there has been some behaviour change on the Google side and the question is if we can code around that or not:
IDENTIFY
, perhaps we bypass the Google cloud entirely?)PROXY_SELECTED
intent is flagging the hub as not available for the proxied reachable devices so Google does not attempt to resolve a local path? I hacked out a ProxySelectedPayload.proxyData to no avail and after not being able to see where the Google Local SDK used it outside of caching for the client, I abandoned it, but might be worth revisiting.Any thoughts are appreciated, I am highly motivated to resolve this issue, it is the last weak cloud link I have in my environment and I am keen to optimise it as much as possible.
The text was updated successfully, but these errors were encountered: