try different java. i had similar issue with Zulu java. after i have switched to OpenJDK the problem is gone.
it could be also that PI3 has not sufficient memory for your openHAB setup. once it cannot allocate more memory it restarts JVM/Java. pay attention to free memory.
It certainly is NOT caused by Java as a root cause so don’t arbitrarily replace that.
Sure it might help as a side effect (e.g. if new Java uses less memory), but you will not know what that changed so you can run into it anytime again.
I’d take a guess at mem usage, too. You might be having a mem leak.
Enable ZRAM if you have not yet done so but start watching RAM and eventually start removing bindings that might be a reason for the leak.
Looking over the span of a minute at 02:15 there are nothing but 3 types of messages:
Z-Wave StartFrames changing
Z-Wave FramesAcknowledged changing
Z-Wave nodes updating
Admittedly, it’s a lot of chatter and, now that I’m thinking about it, 2am is my network heal time. Any reason why that would cause OH to run the startup script though and might I expect some sort of error message if something was overflowing or crashing?
What you are experiencing may be different. When Items or Things are added/removed/updated, several bindings and the old rule engine will restart, which will cause your startup rule to be triggered. This is a known issue…
I suspect that the heal is updating some Things which restarts your rules and possibly the entire zwave binding. I suggest disabling your daily zwave network heal and run it manually when you add or move devices.
I would be willing to bet (and I’m not a gambling man) that this is not the same exact issue as the original post, it has been a really long time since you’ve posted (welcome back) but I’d like to encourage you to start a new thread about your issue Keith, it will help us solve your issue easier
Thank you all for the replies. It seems as though there is something else, unrelated to OPs issue.
I’ll implement this and see if the behavior changes.
There are tons of these around the timestamp of the startup rule executing from my Yamaha stereo and Z-Wave things. To your point, I think if this was the cause, I’d be seeing this rule run every time one of these events occurred.
Upon a closer reading, I believe you are correct. I didn’t realize the consistent frequency of my issue until I started digging into logs for replies. Thanks for the welcome, I’ll go start a new thread.
Not every ThingUpdatedEvent causes any kind of consequence. It depends on the Thing involved, and what changed. zwave Things seem to be culprits sometimes, but you only need one for the effect you’ve seen.
By way of further explanation, these ThingUpdatedEvent are about the properties of the Thing. Not the status like online/offline, which are different events.
So something fundamental changes - I don’t know what, but say a zwave device reports different channels available to what you have configured.
In order for those changes to take effect, some services will need a kick in the pants of some kind. Like when you edit a Thing, and a binding might need to reinitialise it.
At OH 2.5 there seems to be more re-init activity around this than there used to be.
Seeing “tons” of these might be about symptoms, not original trigger.
A recent thread highlighted that a change in zwave binding taken together with a Thing defined under an earlier version could be linked to this kind of problem.
After two month and many time of analyzing / waiting until the error appears again / analyzing again I just wanted to give a short feedback: I identified and solved the problem.
Additionally to openHAB I had one other software installed on the Pi to control my access point. Also every few days I ran an automatical backup on the OH installation.
After analyzing the memory usage I found out: The access point software search for updates every Wednesday and Saturday night. Sometimes (every few weeks) I now had the situation, that the OH backup was still running when the AP software started updating. Because of an memory overflow openHAB froze / restarted.