Hi! I've had no issues for some time on my C-7 hub, but after updating to 2.3.8 (now I'm on 2.3.8.122) it keeps hanging the main processm every few hours.
It is still available on network
The diagnostic tool works
WEB Interface works, but not entirely
Can't access logs at all
No automations work
Dashboards are frozen, even time stops
I've tried to rebuild database
Z-wave is disabled, I've updated it anyway
Once I reboot, it works again for some time, logs show some errors.
Most recent error:
java.lang.OutOfMemoryError: Java heap space (method autorefresh)
the device is Shelly RGBW White, no newer driver available.
Before that the HUB said the load was severe, I've disabled that particular app entirely.
Before that the hub had issue with another driver, which I updated.
All in all the hub hangs on any error, it is no longer reliable.
Sounds like something is chewing up memory. I would check your app and device stats to see what may be hammering it. Also install hubitat device information driver by @thebearmay as it can help you check memory at a glance.
I just had a very similar experience on a C7 hub last night. Very similar symptoms, where some of the hub's web pages would load, but not the logs. Hub had a Sever CPU Usage warning as well. Diagnostics tool worked fine. I was able to reboot the hub using the normal Hub Web UI interface in the Settings menu.
This hub is very lighted loaded with mostly LAN and Cloud integrations. It has no Zigbee, no Z-Wave, and no Matter devices paired with it. CPU, Memory, and CPU Temp shown below.
As you can see above, something consumed a ton of CPU time, causing the Temperature to rise. Free memory was not an issue, as the hub still had ~500MB free at the time of this event.
This hub had been rebooted earlier yesterday morning, via an automated rule that runs weekly at 5:30am. And, I made no changes to this hub yesterday whatsoever. This is a strange one, for sure. I just wish I could have brought up the Logs page to see if I could figure out what was consuming the CPU.
I have been planning on moving the tasks that this hub performs over to my new C8 Pro hub, in order to get everything back to just one Hubitat hub. I have started moving items this morning, like the built-in OpenWeatherMap integration and IFTTT. I will continue the migration today.
It seems that way, but only if there is an error - the app chewing throuh all available memory was a simple shelly integration not being able to connect to one of my shelly devices which was offline at the time...
Thanks! That does look similar in behaviour, can you tell me what the tool is? I'll install it and watch the hub as well,
Thanks! Yes, I've updated this according to the how-to with success and reboot both times as I've had the uit running 2.3.8.122 at the time of uprade. The issues sure did start with 2.3.8.117 so I was hopefull, yet more issues stil arise.
Memory is around 370 - 400 MB now...
Thank you, I've actually tried that, however strangely enough the issues persisted. I went from 2.3.7.145 to 2.3.8.117, then after like a day I saw 2.3.8.118, which didn't help, so then I went back to 2.3.7.145 but to no avail. Now I'm on 2.3.8.122 with the same issues still...
Yes, IP integrations that have poor error handling will quickly tear through all the hub memory hammering the connections.
Sounds like you just have a bunch of poorly made or configured apps and drivers. It probably just all came to a head recently and des not have anything to do with platform updates.
The normal amount of errors in your logs is ZERO. Any regular or repeating error is a concern.
I am using a combination of tools to collect and display the data shown in the above charts. I am using Node-RED to collect the data from my Hubitat hubs using http endpoints. Then, that data is formatted properly and inserted into an InfluxDB server. Finally, I am using Grafana to display the data from InfluxDB. I am running these three tools as Add-Ons on my Home Assistant Yellow hub. They cannot be run directly on a Hubitat hub.
There are other options available as well to achieve the same or similar results. There are multiple threads in the community discussing the collection, storage, and display of HE Hub statistics that you may find interesting.
Thank you! Yes. I agree, however other than platform updates, I've not done anything to the hub.
All the changes in drivers / apps and were done afterwards as a means to resolve the issue...
I'm not a programmer, so I don't write those drivers / apps. The best I can do is keep those drivers and apps up to date...I'm using either internal drivers for Shelly if available, or Shelly released drives, these are now however 4 years old at least...
I've been thinking of upgrading my hub to C-8 Pro now, but untill these are resolved, those errors will just take a bit longer to chew throught the bigger memory...
So are you saying that you have frequent errors in the logs?
If you share them we might be able to tell you what the issue is.... (screenshots of logs please)
If they are system drivers staff cannot fix errors they do not know about. For custom drivers you would want to post about it on the thread where they are posted. The dev or other users may have fixes or solutions.
The Out of Memory error could be a significant problem. Could be that you are polling too often on that device or it could be from something else totally and that driver is just running into the limit.
@DrRomca for a simple way to see your memory history you may want to check out this (new) app:
Only drawback is that it gets reset if you reboot, but it might be useful so you can see how fast your memory is getting low, and if it is gradual or all at once in a big chunk.
