Updated my C8P to 2.5.2.128 this morning and when it rebooted I got a "zigbee channel changed" message from an alert that I set up years ago that's never before triggered. But the channel hasn't actually changed. I've never seen this before. Looking at the logs I'm wondering if the zigbee radio went down while rebooting?
Has anyone else seen this? Makes me leery of rebooting now.
That's scary. My biggest pain is changing ZB channels when I get an interference issue and spending the morning re-associating almost all 70 devices. They almost never find their way home. 
Exactly! The channel didn't actually change and everything is working OK. But it's scary if the radio went down or if the channel actually had changed. What if the radio went down and then came back up on a different channel?
Everything was cruising along like normal with good memory levels, etc. And suddenly the memory dropped from 904 MB to 10 MB, the hub rebooted, the zigbee radio went off (but came back on again). This started after I updated to Zwave-JS, but not sure if that's the cause. I will try shutting down and pulling power. I haven't had it reboot for low memory for ages and other than pulling power I'm not sure exactly how to diagnose this or what my next steps should be. Help?
ETA: I shut down and pulled power, and rebooting did not cause a zigbee shut off. [fingers crossed] Still would like to have some idea of what's happening here.
I don't think it is anything to worry about. It happened during an update, the hub is busy doing a lot of things at reboot, and something got out of sync in boot order for something else taking to long.
Why is it an issue for the radio to shut off? That is a very common troubleshooting step to possibly fix Zigbee mesh issues. Shutting down the hub shuts down the radio after the power drains. Shutting it off for over a half-hour puts the Zigbee devices into panic mode. That doesn't cause any real issues and devices reconnect.
The channel "changed" from "not available" to your original channel, that was the change in the logs, because something didn't update channel in time, or something.
As for the hub memory crash, there are no logs surrounding that posted so you would need to look into it more to find the cause. Main thing is to see if the memory crash happens again or if it was a one off related to the update.
Maybe a @bcopeland thing with your Hub UID?
It's probably not. But I remember a few years ago when my zigbee radio went off and on like a yoyo and things stopped working. So it makes me a bit twitchy.
There's an "Error parsing eventstream data" error from my Hue Bridge about 5 minutes before the memory crashed. But that's never really been an issue and I don't see anything else in the logs other than business as usual.
I agree with @dnickel, the person with the detailed logs and the ability to sort through them is @bcopeland.
Bouncing like that I'm sure will cause problems in the mesh, but that was probably a different issue. How did that problem get resolved?
It seems worthy of being looked into. I've been holding off on Zwave-JS to let it settle, and I have some old GE dimmers and Fibaro devices which could have issues. This update has a lot of JS stuff in it, I noticed.
Speaking of update details, when did we lose the link to the update description on the update page?
With a new hub.
That must be recent because I remember seeing it yesterday I think.
There are a few wonky things, which is to be expected I suppose. For example the local backup defaulted to "full backup," but I actually got both backups last night ā one with the zigbee and zwave radio data, plus one without. Maybe because the full backup is still in beta? Or maybe just something weird?
It was there today when I updated to 128.
Now you have me imagining things. 
Just to follow up, there have been no new issues. Memory seems to be stable. I have a reboot scheduled for the 1st day of the month. It rebooted last night and the zigbee radio stayed online. Perhaps the gremlins have been banished. Thanks for listening to me whine.
