As per the title, I have installed and re-installed HASSIO over 10 times now and I’m guessing this is a rare/unique occurrence…
At first, usually for the first 1-2 weeks all of my components run just fine… I have a mix of Wifi and Xiaomi (zigbee) products. Then somewhere between 2-4 weeks without any change to the config things get … buggy. Usually it’s the zigbee devices becoming unresponsive, initially within a 24 hr period of being rebooted, then this timeframe narrows to about 30 minutes! The errors are always Gateway Hub related (No Response from the Hub or Invalid Key), accompanied by a whole stack of ‘Failed to Connect to Bulb’ issues.
I have reinstalled it on a different SD card but I get the same issue which makes me wonder if it’s the PI itself … or that Xiaomi zigbee products are just not great.
Has anyone experienced this issue with HASSIO? I love HA but this is really killing me
I switched from the All-in-One installer on a Pi3 to Hass.io last fall and find that it is more stable and easier to update. I have never had to reinstall it since moving to Hass.io. The difference here being that I use zwave, not zigbee, and have no hubs or cloud service dependancies.
I was under the impression that Hass.io was the all-in-one installer for PI…?
I was wondering if moving away from the xiaomi stuff would yield better results… I’m just hesitant since if xiaomi were that problematic I would have thought there would be a lot more noise here about it.
Nope, Hass.io is a docker based implementation using resin.io as the ‘core’ or boot os.
The AIO installer was an older (now deprecated) build that built a virtual environment, and openzwave on top of an existing OS, such as Raspian.
Have been running Xiaomi devices ( 26 switches and sensors) for almost a year , first on Domoticz now on Home assistant. Never had an an issue with Xiaomi kit. I have had problems with Hass.io but “touch wood” since switching from WIFI to Ethernet it has been stable.
I can confirm as part of my testing when Hass.io has been unresponsive , the native Xiaomi app works perfectly . In all this time I think I have rebooted the Xiaomi gateway once.
I’ve been running 15 xiaomi zigbee devices for 5 months now and have never had to reboot the xiaomi hub. Everything works extremely reliably and the only issue was a temperature sensor that was near the range limit and moving it a bit nearer to the hub fixed that.
From the HASSIO side, I’m connected by Ethernet and have never needed to reboot. I do get occasional “no response from hub” messages in the log but this has not resulted in losing any of the devices in HA. I suspect those messages are more likely connected to what load is on my router at those times.
When the devices become unresponsive, are they still visible in the xiaomi app? If so, you can narrow the problem down to HASSIO. If not, it’s on the xiaomi hardware side and you can look to see if there are range or battery level issues. If it’s HASSIO, you could try Ethernet if you are currently using WiFi?
Everything works perfectly via the app, and yes I’ve been using ethernet from day one.
I specifically ordered a power supply for the PI having read about power issues…
I’m thinking of installing HA on linux (python / docker) on my esxi machine to see if I get the same issues… not looking forward to it, but I’m at my wits end
I was just thinking the same thing, I have not touched my install in a while and when I went to load it, there was some error. Wondering is Pi is just not a good enough for this type of always on application.
I have had very few real issues with stability. I use a spare Apple branded usb charger from a phone, and a good solid Samsung Evo SD card along with Ethernet as opposed to Wifi. My move to HassIO has made things so much easier.
I had my configuration working perfectly fine, I was able to connect, all my automation’s were working and then the next day without doing anything…poof, unable to connect. I don’t think I have had this running for more that 1 week without some issues. This is connected via LAN with the latest.
