2021.9.0: More energy, USB discovery, template ❤️

I have a really bad idea…

If you have a sensor for how much solar you are generating and sensors for how much is going to the house and how much to the grid you could make a fake sensor.

Production minus export minus consumption = going in to battery

I updated some lights using tasmota and errors are gone. But the supervisor not starting anyway.
It was trying to update from 2021.8.8 to 2021.9.2.
I’m still on 2021.8.8 but with no supervisor, so no command line. I use docker but not a fluent user.
HA is functional but I got stuck.

PS: In Docker, I have “hassio_supervisor” “running”…, but in HA I have:

Acessing the Observer (port 4357):
image
:woozy_face:

It seems that supervisor got updated to 2021.9.0 but hassio is still in 2021.8.8 (from the docker supervisor container:

There are, the thread will confirm.

@MartinB @jivesinger @gulli
Think I have fixed the issue with energy not calculating costs anymore after the update.

EDIT: WARNING - MY DB DIED the next day with errors;

Error executing query: (sqlite3.DatabaseError) database disk image is malformed [SQL: DELETE FROM events WHERE events.event_id IN 

Unrecoverable sqlite3 database corruption detected: (sqlite3.DatabaseError) database disk image is malformed [SQL: DELETE FROM events WHERE events.event_id IN

The system will rename the corrupt database file //config/home-assistant_v2.db to //config/home-assistant_v2.db.corrupt.2021-09-06T03:12:48.885687+00:00 in order to allow startup to proceed

I think we need to stop the DB first, I am not 100% this is what caused it but just wanted to make sure people know there is a risk.

Bug was here

I installed the SQLifte Web addon and checked, the DB was showing EUR but the sensor showing as GBP so the error about them not matching makes sense.

I installed DB Browser for SQLite and checked against the DB which I connected to using the samba share plugin and edited the DB to GBP to match my sensor then rebooted.

The errors have gone from the log and my sensor seems to be updating again now;

image

Will need to check once we get 12 mins past the hour when the energy panel updates.

1 Like

I have had continual problems with supervisor not starting after it updates. I do just then run the install script but have never needed to restore a backup/snapshot…

Again I did not say that it will not work on a Pi3. What I said is that most of the beta testers have moved on from the PI3 as it is now considered underpowered. Lots of write ups in the forums, reddit, etc saying the minimum for HA is a Pi4 4G . The tone above was blaming the developers for not testing the new release on the Pi3. Others pointed out that as this is open source, volunteer to beta test or wait until .1 or .2 is released.

I have seen lots of issues of the years of using HA, but there were always people here, on discord, Reddit, etc and developers that would help resolve any of my issues fairly quickly (whether caused by me (most of the time) or a software bug in HA.

The Installation documentation states an RPi 3 is “OK too”.

Based on collected analytics, the RPi 3 represents about 13% of all host systems, in third place behind RPi 4 and VMs.

2 Likes

It’s still pretty obvious there are not many/any people using the beta on a RPi3 hence the issue slipped through. The solution is more people using that hardware being involved in Testing… if you can’t or won’t step up then… (not you Taras)

I agree with you that given the bug’s severity, it would suggest there were precious few beta-testers on RPi 3.

I don’t participate in beta-testing so I’m in no position to comment on those who do (other than thanks for your contribution, large or small).

1 Like

There does seem to be a prevailing attitude with some people that ‘someone should have tested this damnit’… but that someone is YOU…

7 Likes

Thanks for the info.
I probably shouldn’t have made DB changes while HA was running.
Let’s see if I’ve killed it… :grimacing:

I have tried twice to upgrade to 2021.9.1 then 2021.9.2 using the following commands. sudo su -s /bin/bash homeassistant, source /srv/homeassistant/bin/activate, pip3 install --upgrade homeassistant. All of the outputs say the installation completed but when I try to connect with the browser (http://192.168.1.17:8123/lovelace-yaml/default_view) I never connect. When I reinstall 2021.8.8 with pip3 install homeassistant==2021.8.8 every thing works. Any help would be appreciated.

You need to read the startup messages and see if there are any errors.

Since the new update something in the yeelight section is broken! It seems they are not syncing exactly anymore. My Lamps show me that the nightlight/moonlight is on even the lamp is off.

I downgraded back to 2021.8.5 and it worked perfectly again

To clarify: in the states “night_light:” its always true and so the binary sensor :confused:
I can deactivate it manually in the yeelight app but if i use the moonlight mode again i’m in a loop!

And yet you present no logs or useful information?

I did the same thing, I have numbers this morning.

And I haven’t got any errors in the logs, so hopefully all good.

I have also integrated my E3DC via Morbus integration into home assistant.
To use the provided information in the new Energy dashboard you need to create some additional sensors specifically for grid import/export and battery charge/discharge like it is specified in this thread.
Following you must use the Riemann integration to convert the values from power to energy.
However at the moment I am not fully satisfied with the accuracy, of the provided values. This may come due to the different pullrates, but I have not further investigated so far. The official E3DC is updated only every 15 minutes, where as the Modbus integration pulls the values every minute.
The values visible at the E3DC Webinterface Match the values on my meter perfectly.

With 2021.9.3 my Shelly is back :grinning: Thanks Team.

Hi, since 2021.9.x my interactive floorplan reloads every 2 to 5 seconds (and getting those blue waiting circles on every button for a second.

Does anyone has a solution for that?

Do you have any messages in your log, maybe check the logbook as well see if devices are going online/offline