Evohome app broken

[anniesboy]

Odd but this site
https://downforeveryoneorjustme.com/

Suggests it is not down,even though it is

[adamski]

Definitely still down for me.
I had to actually go and change the thermostat by hand this morning, Alexa would not obey my command!

[FullBore]

I have just received a response from support:-

Good morning,

Thank you for contacting Resideo-Honeywell Consumer Support. Your case reference number is nnnn471.

Unfortunately we have some issue with the servers . We are aware of this situations and we are working to correct the issues as soon as possible.

Please use the system manually and try again to use the app starting with the afternoon. (sic)

Thank you very much for your collaboration.

Kind regards

Rather disappointing that it appears that the servers won't be available until this afternoon.

NJSS

[gordonb3]

Rather disappointing that it appears that the servers won't be available until this afternoon.

The servers are all US based, so in Europe you're at least 5 hours (UK, IE, PT, IS) ahead of the technician waking up.

Odd but this site
https://downforeveryoneorjustme.com/

Suggests it is not down,even though it is

That likely depends on your definition of `down`. The server is in fact responding. The problem is that it only returns a `HTTP 503 : service unavailable` error.

Interesting fact: there appear to be two(!) different servers allowing access to at least the older API and they both have the same issue.

That's a bit hard to judge as to me it seems that the heating system uses a different type of communication to the Total Connect servers than the user app.

So I figured I might just as well verify that statement and it turns out that the RFG100 does in fact access a different range of servers for exchanging information. Surprisingly much I should add.

[garymtitley]

Yes your right,and it’s still not working!!!

[DBMandrake]

Have any of you Domoticz users ever collated the down time? I'd just be interested to know what the service level is. I'd guess around 99.25%.

I don't use Domoticz anymore, however I do use evohome-munin to graph my temperatures, which uses the same V1 API.

90% of the time when there is an outage that affects the mobile apps, my graphs are not interrupted in any way, and I very rarely see any interruptions in my graphs. (At least recently)

From this I have to conclude that in the majority of outages, the upstream link from the Evotouch to Honeywell is OK, and the old V1 API is working OK, however the V2 API which the smart phone apps use is down.

However I checked just now and my Evohome graphs using the V1 API are completely dead with no data since 12:50am and were working perfectly without any dropouts right up until then. I have never seen this happen before so I think something big is afoot.

I actually noticed the smartphone app wasn't working much earlier than that - from memory about 10:30pm I tried to turn off the downstairs heating with my usual "Alexa Goodnight" routine, this failed to work and sent me a bunch of zone change failed emails, I then checked the phone app and was greeted with the "down for maintenance" page which I still get now. So I think it may have been partially down (V2 API) for maintenance as early as late evening but everything went down by 12:50am.

On the V1 API I get:

"Connection failed: 503 Server Error: Service Unavailable for url: https://tccna.honeywell.com/WebAPI/api/locations?userId=xxxxxxxx&allData=True"

and on the V2 API I get:

"requests.exceptions.HTTPError: 503 Server Error: Service Unavailable for url: https://tccna.honeywell.com/Auth/OAuth/Token"

Doesn't look good!

[frankmalia]

APP is down, system working.

Yeah my graphs are returning the same data as 1 am this morning.. so they look great and steady

IFTTT did not work yesterday on geofencing either..

[DBMandrake]

Yeah my graphs are returning the same data as 1 am this morning.. so they look great and steady

My scripts recognise a lack of fresh data and put an actual gap in the graph instead of just repeating old data.

[gordonb3]

From this I have to conclude that in the majority of outages, the upstream link from the Evotouch to Honeywell is OK, and the old V1 API is working OK, however the V2 API which the smart phone apps use is down.

It's actually mostly the 0Auth method that fails. I rarely see errors from the actual API. Depending on the source of 0Auth failing this may actually become a problem for the v1 API as well since Honeywell stripped session renewal from it.

[FullBore]

Hurrah

It's back again at 16:25 BST.

FB