I'm a bit like you. Ipad, iphone (because they just work) but for my main PC I use a Dell XPS 13 and a WD19s docking station. It stays up for weeks (even bouncing from sleep mode which is usually trash on most systems). I reboot for the occasional update is about it.
Oh very good with that threshold thanks. I have mine set way to high right now, will adjust.
Good to know, thank you.
I just dialed mine down to 100k for both hubs, should be interesting to see how it goes.
Great to know. Thanks for the info. Will adjust my reboot threshold as well (more conservatively tho at 125,000kb).
HOLY !! I haven’t seen mine drop below 270; and when it dropped to that level it caused me anxiety (no noticeable problems with the hub, just with me). I set my warning to 250, which is down from 290. Just a little gun shy from prior experience.
My main C7 is down to 295k and still working well - I set the reboot threshold to 100k just to see what happens.
Mine usually take a week to drop that low, and then seem to plateau around 280-290, but I’ve been playing around with things lately and rebooting more frequently because of it, so I’m not sure if that’s still true.
My Primary C7 does all the heavy lifting in my house, my Secondary C7 only runs cloud integrations and it runs for a lot longer in-between reboots.
I’ve tried to keep my 2 pretty evenly stressed. My database sizes are small due to decreasing event and state history size as much as possible, as well as event/state days. I also don’t use Maker api. I have over 200 devices on each hub.
I still have a few cloud integrations that I've been too lazy to move over - plus my main C7 handles both zigbee and z-wave, plus 98% of my automation.
I understand. It’s been a few years since I added the second hub and redoing automations is tedious work.
The recently(?) added feature of export/import has really helped me in moving automations from one hub to another. Especially if all your devices will have the same names, it's easy peasy.
So nearly 5 days has passed since my last reboot and the hub is still running well,. Even tho free ram is down to 263k.
That is also a great "time saver" type of app, which should make writing apps (that are very similar to existing apps) much easier.
However, in my case, I've established a number of apps that I want to use on each installation that I do. (With perhaps some minor modifications). In that type of case, exporting those apps to an external file, and then importing them, becomes a simple, easy to use way.
Although "Automations" do not have a simple "wsiwyg" interface, nonetheless, it can be relatively easy to use. This is just one aspect of the overall "Automations" in Hubitat. This feature and those like it are the reason that in my humble opinion, it's much more advanced (in Automations) than it's competitors.
So I’m now 8 days in and under 250k free memory and my c7 is still running perfectly.
I don’t know what the Hubitat team did in 2.3.x, but I’m impressed with the improvement!
Hmm, so the further under 250k I got, the less stable my Zigbee mesh got. Nothing else (eg zwave or automations) was impacted at all, but I had 3 devices that needed to be re-paired, 3 mains plug sockets and both my zen thermostats needed to have their batteries reinstalled to start responding again.
FWIW I have a high quality external Zigbee antenna.
So what is the consensus? Should I raise my reboot threshold again despite suggestion from the HE staff? Any revised recommendations? TIA for any input.
i would check your mesh .. my memory is always below 250 and i have no such issues.
I've got a pretty decent size mesh (23 devices, 2/3rds are Mains powered) and an even bigger Z-wave mesh (more than double the devices). I also have a 5.5 dBi Rubber Duck Zigbee Antenna upgrade installed.
My hub was working well under 250k free ram, except for my Zigbee devices started playing up and some even dropped off the mesh or just stopped responding. Everything else was operating 100% normally. After a reboot my Zigbee issues went away.
Anyway, I've set my reboot to 250k as that is around 7 days of up-time between reboots and imo quite acceptable.