I have read numerous posts about random HA crashes due to SD cards. I was having to re-cycle power several times a day on my pi due to crashes. I just upgraded to the new HA Blue and it ran fine for about 2 months, but now it’s crashing several times a day. I can’t find anything in the logs that would point to a reason. Can anyone offer any help? Let me know what you need to see to help diagnose. It’s kind of worthless to have HA if it’s going to be crashed every time an automation is supposed to run.
I’m having the SAME thing. In the last month the time between crashes has gone down. Now it’s like every 8 hours it locks up.
I have noticed the memory usage rises over time. I’m assuming some type of leak but I don’t know how to find it.
I installed the system monitoring feature and nothing ever shows as a spike there, although there’s no way to see the history after a crash so who knows. I have plenty of free memory and disk space and even my swap only hovers around 30% usage. I had thought it could be a buggy card I installed via HACS because when I took it out things ran perfectly for a week. Then yesterday it started crashing again and the only thing I had done differently was install the Xbox integration. I went ahead and took that out to no avail.
‘’21-07-25 15:53:04 WARNING (MainThread) [supervisor.addons.validate] Add-on config ‘devices’ use a deprecated format, the new format uses a list of paths only. Please report this to the maintainer of picamera
21-07-25 15:53:04 WARNING (MainThread) [supervisor.addons.validate] Add-on config ‘devices’ use a deprecated format, the new format uses a list of paths only. Please report this to the maintainer of Tailscale
21-07-25 15:53:04 WARNING (MainThread) [supervisor.addons.validate] Add-on config ‘devices’ use a deprecated format, the new format uses a list of paths only. Please report this to the maintainer of pilight’’
What is this in my Supervisor log?
I would expect you need to let the maintainers of those custom addons know that there’s issues with them. The error may not make sense but it at least tells you what to do
That would make sense only I have no idea what “integrations” those are associated with.
They are not integrations. They are add-ons. Each of these addons will have a link to their repository in the Supervisor Dashboard.
picamera
Tailscale
pilight
I have no add-ons with those names. How do I trace which they belong to?
That presents a dilemma because the Supervisor believes you do.
Click this:
Are there any addons liste on that page?
In the Supervisor Addon Store, do you have any custom repositories listed?
Several, but none with any permutations of those names, i.e. picamera, Tailscale or pilight
That’s where it is coming from. The custom repositories contain those addons and even though you have not installed the addons they still generate the warnings. You can ignore them. They are not causing your issue.
I had a discussion with Frenk about this causing confusion a while ago:
2 Likes
Out of curiosity, what should my Core RAM usage be? I find mine hovering between 61 and 65% and am wondering if this might be a cause. Or even more so, my overall memory usage. Just saw a ton of spikes.
That’s quite a bit. (60% of 4GB = 2.4GB). Mine is sitting around 5% of 8GB = 400MB.
Have a look at how much Node Red is using.
According to Supervisor, 2.8%
The Profiler integration may assist in finding out what is consuming all your resources.
Installed, but not sure how to extract info
When the profile is complete, Profiler will generate a Python
cprofand acallgrind.outfile in your configuration directory.
You then need to use one of the utilities listed in he docs to open the files.