1min sensors not updating (but not all?)

[ericwages]

It appears that 1min devices are not being updated consistently. Some of them look like they track well, others only update infrequently leading to very sporadic information.

First, a picture of what HA is seeing:

What Emporia is seeing:

Reloading the integration does not solve the issue.

[magico13]

There are a few issues tracking this. The live data from the API is unstable, which is what the integration pulls from. The history in the Emporia app is compiled from past data when requested and has the benefit of being able to work with delayed data.

The only thing that's a fair comparison to make between the two is the main page in the app, set to the 1 minute level, and the "now" option. If the data is messed up when viewing the app with those settings, the integration will be too. Unfortunately that data is often bad in the app.

[ericwages]

I can confirm that the 1 min view on the web and android app versions of Emporia are both showing power data whereas the 1min on the HA has not updated in 5 hours.

[joshs85]

All of the 1M sensors become unavailable after some time for me.. and a reload of the plugin seems to resolve it.

[dabbler68]

Same here. Reload fixes it. Reboot makes it fail again. So a workaround is to remember to reload after rebooting, I have not seen if reboot is the only trigger.

The message in Logfile is:
2023-09-23 21:53:13.193 WARNING (MainThread) [homeassistant.components.group.sensor] Unable to use state. Only numerical states are supported, entity sensor.oven_7_1d with value unavailable excluded from calculation

Strangely, one sensor is not reported in the logfile but, like the others in the list of entities it is marked as unavailable and does not update. The Day and Month sensors are all OK.

Mine was only set up 2 days ago I can only speak for the current 2023.09.2 version of HA.

[ericwages]

What I see that a reload does not necessarily fix ALL the sensors. 1m sensors for the major devices (the hubs) are always reliable but the circuit sensors are intermittent. I usually have to redownload to try to get it back up, but even then, it's not reliable for long. I have two emporias - one on the main panel and a second installed on a sub panel. (It's configure as such in the tool as well)

…

On Sat, Sep 23, 2023, 08:00 dabbler68 ***@***.***> wrote: Same here. Reload fixes it. Reboot makes it fail again. So a workaround is to remember to reload after rebooting, I have not seen if reboot is the only trigger. Mine was only set up 2 days ago I can only speak for the current 2023.09.2 version of HA. — Reply to this email directly, view it on GitHub <#219 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A3ZC4G4LLPRG6P2BS63WCFTX33FNZANCNFSM6AAAAAA3E3CQYU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[dabbler68]

Today I had a look at the logfile and there was a difference.
Up until 04:16am yesterdaythe "Unable to use state ...." messages had been coming at irregular times. But then at 04:17, those messages were replaced with:
2023-09-26 04:17:29.427 WARNING (MainThread) [custom_components.emporia_vue] More than 30 seconds have passed between the requested datetime and the returned datetime. Requested: 2023-09-25 18:17:29.160768+00:00 Returned: 2023-09-25 18:16:59+00:00
These continued approximately every minute with the occasional "Unable ...." message as well, After about 30 hours they stopped of their own accord. But the feed to HA continued all the time.

I am guessing it may be related to Daylight Savings time in some country and was fixed by Emporia. (DST has not started here yet ... Australia).

As for all sensors or not I am seeing some consistency, I have 8 sensors. Sensor 6 has never reported the "Unable ..." message while sensors 4 and 5 always report twice in the same group of warnings. All sensors are reporting their energy usage.

When the integration is reloaded, the 1 day sensors also report the "Unable .. " message. But for these sensors 4 and 5 do not duplicate,

Maybe as there is somce consistency in these messages there is something in these notes that will be meaningful when looking for a solution.

[CAHutchins]

I'm losing all of my 1Min sensors after a reboot. And occasionally just randomly when it's been running fine for a while.
After a full reboot all the 1Min sensors all go "Unavailable" until I restart the integration.
I have two Vue2s with 20 branch circuits in total. I wasn't having this problem before when I only had a utility connect so it might be related to the number of sensors I have now.