bmercier

@bpwwer, is it that @automationgeek no longer maintains this plugin?

Chris owns these 2 tickets. I asked him to review them on Monday. Sorry about the delays.

Hi, First of all, I would not call it a "database failure". The database itself was never corrupted or had loss of any data, which a "database failure" would imply. I'm not sure of all that you did, but all that was needed is to relink. To answer your question, no there is no upload function. The download of the spoken exists to keep a reference, or to create documentation on the spokens. It is not meant to be a backup process. I'm not sure if you deleted them or not, but you should not. Now regarding your program spoken - Are you saying that when you edit existing spokens that are programs, the Spoken and Alternate Spokens fields are empty? I'm not sure why that would be. I looked up your account, and edited the same one as in your screenshot "Siren-1second (0065)", and I do see the spokens. I see it for all of your programs. I'm thinking that you might have a browser extension that fills these fields, actually clearing them with an empty string.

@BEdgar Whenever you need support, please open a ticket. We don't actively monitor the forum. There were other threads related to Google Home not working, and as far as I can tell, everyone has their Alexa or Google Home working. The Google Home issue is not related to any changes in the eisy/polisy or Portal. The problem has been explained here: So back to your Google Home problen, after linking, even though it does not show as Linked, it probably is Linked. Can you check if you can see the status of devices and control them? Could you create one more spoken and click "Send To Google"? I had that problem, but creating a spoken and sending it to Google made it appear as Linked. One more thing to try, I would suggest to Delete your Home and create a new one. Go to Settings | Delete this Home There is something wrong in your Google account, something needs to be reset. By experimenting, the "Send to Google" should be able to fix it.

Yes I've seen that. But my question is - Were you able to reproduce the problem?

@Panda88 Were you able to reproduce the problem?

@johnstonf Even though it does not show as Linked, it probably is. Can you check if you can see the status of devices and control them? Could you create one more spoken and click "Send To Google"? One more thing to try, I would suggest to Delete your Home and create a new one. Go to Settings | Delete this Home Something needs to be cleared in your account.

Then I have no clue why Google would not show it as Linked. That's not something we can control.

Why would you say that? As I explained, this was cause. Google could not refresh the access token during the outage. The refresh token Google has is no longer valid, so you have to re-link. Nothing else changed.

@johnstonf, I'm sorry that you are still experiencing issues with the Google smart home integration. I don't know why it would not show as Linked after Linking. That would be a question to ask Google support. This kind of trouble can happen when using multiple Google accounts, so perhaps it used to be linked with a different google user id than what you are using now? That's the only thing that comes to mind. If that can help, there has been no changes to the Google integration in a long time. The only thing that the outage caused is that Google could not refresh the access tokens if it happened to be due for refresh during that period. If the google smart home integration does not show as Linked after linking, there is not much we can do to fix it.

So here's what happened. There were 2 bugs in the code of the proxy servers that were dormant. They had no effects, unless a proxy server had a hard stop... which happened last night. The first bug had the effect that some units were trying to connect to a no longer existing proxy server. The second bug is more complicated to explain, but it was also caused by the hard stop. This one was causing the HTTP 500. The hard stop was caused by AWS stopping one of our proxy instance unexpectedly. This was NOT a graceful shutdown. This is usually caused by the failure of the underlying hardware. Both bugs have been fixed and are in production.

This should be fixed for everyone now. Currently doing additional checks. Will post a summary later.

Should be working now

@Sir @hum099,Can you please try a reboot?

Currently investigating. Some units are having trouble with Portal.