Also posted on subreddit. Key factor for me is there were no warning indicators such as severe hub load.
Last night at about 4:30 Am (according to my Hubitat log) my hub became unresponsive (crashed?). I became aware of it when none of my house wake-up rules and room lights instances ran. After verifying that Lan, WiFi and internet were all operational, I just rebooted hubitat by unplugging power and then plugging in again. After that, everything is good again.
Coincidently, I had updated to the latest hubitat firmware version last evening (C-7, 2.3.4.134). Other than that there have not been any changes to my devices, rule, etc. in quite some time.
This is the first time I've ever experienced what appears to be a hub crash. The hub has always been rock solid.
Incidentally, there were no warnings of unusual hub load. I don't know if there is a way to check memory usage and processor load to look for signs of being over taxed, but I doubt that is the issue unless something like memory leaks can occur.
So my questions are:
Is it just a coincidence this happened after upgrading firmware?
Has anyone experienced anything similar, now or in the past?
Any comments, insights?
Thanks. Baffled, but not too concerned at this point...
Before you unplug power you should go to settings and use the shutdown button. Once the hub led turns red then you can unplug and wait a minute or so. Uncontrolled unplugging can cause database corruption.
Always unplug from the outlet, not from the hub, which could cause a weak connection. You can also use the reboot button from the settings page.
To add to @Sakman’s recommendation, if the main Hubitat user interface isn’t working, going to http://hub.ip/8081 might sometimes work and also offers shut down and reboot options. Generally a good idea to try these as pulling the plug without first shutting down could result in a corrupted database. (Restoring the last backup will resolve this however).
Was there anything in the logs pointing to errors?
I have a rule machine rule set up using the information driver to auto reboot the hub when memory usage is super high.
This used to happen to me all the time. The culprit was a misconfigured circadian daylight instance that spammed commands because I had two different instances fighting over the same bulbs.
Very good to know. I couldn't reach the hub via the UI. It looks like nothing corrupted but I always back up after updates or changes I make.
No the logs looked completely normal with no errors or signs of problems, just some periodic reporting of lux values from philips hue motion sensors and battery level from a U-tek door lock. All had been fully quiet (automation wise) for 5 hours. I'm not sure why these devices report status. I think it is a u-tec firmware problem. The lux is supposed to report on a 10 lux change, but all were reporting 0 since the house was dark.
