Open jr298 opened 1 week ago
Update: Installed tcpdump on home assistant, and can't see any traffic to the inverter IP.
This looks very similar to the recent AIO issues (#189), but this isn't an AIO - and it's been running for months. Any suggestions for further diagnostics?
From my own experience of having the same issue occur on my Gen 1 hybrid, and of others that have randomly had it happen, 4 suggestions:
No idea why it happens either
I tried deleting the PKL files - nothing Tried deleting and re-installing GivTCP - nothing Reset Inverter - that did the trick.
What the.... At least all my automations can carry on now.
Also - Thank you!
Now - I wonder if I can detect the issue in home assistant and at least send myself a notification.
Well - looks like the best indicator will be "Last Updated" So I've added a helper which will update when that updates and every minute:
{{ as_timestamp(now()) - as_timestamp(states.sensor.givtcp_ce22*****_last_updated_time['state']) }}
And now I have an automation which notifies me if that exceeds 100 seconds.
Well - looks like the best indicator will be "Last Updated"
So I've added a helper which will update when that updates and every minute:
{{ as_timestamp(now()) - as_timestamp(states.sensor.givtcp_ce22*****_last_updated_time['state']) }}
And now I have an automation which notifies me if that exceeds 100 seconds.
I have an automation that monitors this and a bunch of other things like battery temperature or add-on not running as indicators of failure that you are welcome to reuse: https://springfall2008.github.io/batpred/output-data/#givtcp-activity-monitor
I have the 'staleness' test set to 15 minutes not 100 seconds as occasionally you get inverter comms issues that would give false alerts otherwise
I noticed this morning that the last updated time just stayed at "early morning" until I restarted GivTCP, then it went unknown. I'm hoping that my helper will detect this scenario - I was alerted to an issue by my "peak export" notification - obviously too late to do anything to get the battery to charge.
Where and how do you toggle old firmware to false? I've seen people suggesting this but can't find anything relating to old firmware in any documentation. Literally have no idea if they're talking about the docker stack or something else.
'old firmware' is in the add-on config in Home Assistant.
If you're running GivTCP directly under docker, sorry can't help. Would imagine it'd be in a config file ??
Having been working reliably for months (thank you) GivTCP suddenly failed to talk to the inverter last night - it looks like this was just after midnight.
I've therefore:
Add-on is currently at 2.4.3 Talking with a Gen1 AC coupled inverter with one battery. Last inverter firmware update was in December (Inverter:
D0.535-A0.535
; Battery:3015
)This looks very similar to the recent AIO issues (https://github.com/britkat1980/giv_tcp/issues/189), but this isn't an AIO - and it's been running for months. Any suggestions for further diagnostics?
Cheers,
MQTT explorer shoes me that there is a current*
GivEnergy/Serial/status=online
message, but nothing else in the GE tree.I don't always get the traceback, but there is still no connection: