I have been running Hass.io (HassOS 4.12, Supervisor 234 & core 1.114.1) for a while on a Odroid N2 with a 64Gb eMMC. A few weeks ago I changed from the normal sqlite database to the MariaDB addon. Hardware wise I have a usb RS485 adapter included.
The configuration includes two custom components, one for reading modbus TCP solaredge inverter data and one to crawl the solax cloud for my secondary inverter data. The mariaDB addon has a 2,5Gb database, the recorder keeps a month worth of data without any filtering on usefull/less sensors (this is still a todo).
A week ago I experienced a crash on the Odroid N2 where it was impossible to boot again after reset. To ensure there wasnt any hardware problem I ordered the new N2+. I restored the backup and everything was good for another 12 hours. Then the same issue occurred. I have tried reading logs to see what the problem is but my feeling is that the normal logs aren’t persistent.
At the moment the installation needs 2/3 hardware resets per day to continue. The latest freeze I looked at the console for any errors but I couldn’t find any. Now I have top running in the console hoping to find the issue on the screen at the next freeze.
Do you have any suggestions what I can do to prepare my system to find the issues in the logs or perhaps changes in configuration that might help?
Hi, I am experiencing the exact same problem. I upgraded from the XU4 to the N2+ recently and having hard crashes every few hours. I reverted back to the XU4 for now but there seems to be a problem with Hassio and N2+. I installed another OS on the N2+ and no crashes so far. Not a hardware defect.
I have tried to disable various parts of hassio that might impact the system. My 50 heatpump sensors over modbus that are polled every 10 seconds have been disabled. I suspected the USB serial modbus adapter. That disabled still caused crashes.
I suspected a custom tcp modbus component for polling my solaredge inverter. No crashes today with overcast. Still researching that possibility.
If there are however more ppl with this issue we might try and bring this to the hard kernel forum?
I did a restore to a RPi 4. My crash probe setup with the N2 and recently the N2+ is very stable. If only there would be faster and better I/O resistant storage options for the RPi.
Hope someone finds a solution or a patch for the odroids
My RPi4 setup has been stable for a week, now I also tried the same config with the vmware image. This is running stable for 3 days.
Since the initial crashes on the odroid my database on Maria has grown with 2gb to 4. So I’ll probably stay on vmware for now. This has a good mix of speed, disk io and fits in an easy backup strategy. I’ll see where I can use my odroid n2 and n2+
This week I tried updating my N+ to see how it performed under the latest OS and Core updates. It isnt crashing anymore. However I’m not running my production workload on that device anymore due to my current vmware setup.
I have recently set up a clean HA OS install on an Odroid N2+. It gets stuck about every 48 hours. I cannot reach the HA interface and the blue processor light on the Odroid stops blinking. Can someone tell me how I can ensure that the logs are saved, even if I have to restart the device? This will help with analysing the cause of the crash.
Hi, did you manage to solve it? HA blue just arrived fresh out of the box. The system is reachable for 20 to 30 minutes and then only a power cycle works after which the problem repeats itself.
My Odroid-N2 recently started to show this behavior. I cant figure out what the problem is, at first I thought it had something to do with the daylight bug that was recently discovered.
Upgrading to its newest version did do the trick for a week or so, but now the issue is back.
Running the folllowing versions:
core-2021.11.4, restored a back-up with this version to see if anything is different.
supervisor-2021.10.8, latest.
Home Assistant OS 6.6
When it crashes down, I can connect with putty to the Odroid but logging in with my credentials gives an access denied (works when its up and running). So thats kind of strange.
I can access my web interface of the unify controller that is running on the Odroid.
If anyone could help us all out of point us into a direction that would help it would really be appreciated.
Restored a backup, waited a few days and everything is running smootly. After a few days I upgraded to the newest versions and everything is still running like it should.
The one thing that I didnt add to it is the possibility to cast lovelace to my Google Nest. Still having the cast integration but not using it.
I’m currently running 4.17 of the OS, that seems to resolve the crashing.
Everything is running stable for a few days now.
After upgrading to 6.6, my odroid crashes and is unresponsive, even via ssh.
I’ve done some research and it could be a kernel issue from what I’ve read.
But I lack the skills to really look further into this.
Hi, i also have the same problem on odroid n2+. But decided to do a clean install. Now problem seemed to have happened only when i updated the home assistant. At first it was updating as normal. But then suddenly the blue light stopped. Indicating that it’s stalled. Checked on that power monitor plug on the adaptor, the watt goes down to 1 watt something. Normally around 2 to 3 watts.
Then i plug unplug, it goes normally into home assistant, but still not updated. Then i updated again, and it works. This happens twice already between 2 recent home assistant updates.
And also sometimes there are times where it just randomly stall/hangs. No blue light indicator. But this occurs on roughly once a week.
Any idea guys?
Had another ha installation on parents house running on raspi 4, never was any problem of stall/hangs. Which is quite evious. Always update whenever a new updates available.