Same here. on 134 and the issue has returned.
Platform Version
2.3.4.134
Hardware Version
Rev C-5
Solo unit - was having severe load issues until I used the reboot option on the settings page of the hubitat. Since then - 4 hours so far and no severe load warnings.
where can i see all these cpu stats people are looking at?
I seem to be seeing the severe load when available memory drops below about 240kb - if I reboot, then things are fine again:
Today's peak:
The load of 3.39 is the issue. That could be a high number of apps/drivers waiting on responses to calls - generally internet integrations, or a runaway task/app/driver.
1 Like
i wouldnt call 40% calm. i have 300 devices and many cloud and have never seen that high.
hopefully they will find the cause soon.. it doesnt appear to be a device or app as none show that high.. seems like some kind of underlying os thread but not everyone has it so it must be caused by some combo of devices or apps.. Very hard to track down..
2 Likes
Yeah, it's not ideal, I've only got 145 devices on my primary hub.
It's definitely still an issue too - this morning I knocked up a Severe Load management rule, and it's triggered six times before lunch. I think I'll just set it to kick the hub over when it gets to 10 alerts, regardless of the time of day.
I spoke too soon, the high CPU issue still exists, I have a rule that reboots the hubs if free memory drops below 225 MB and the hub rebooted after 37 hrs of uptime.
225 may be too low mine both go below 200 and still work fine. I thinl the threshold is lower with the addition of the homekit.
Keep an.eye.on.it but i wouldnt auto reboot.
I would, I have mine reboot when it's below 225 free at 4am, If I dont, performance starts degrading rapidly.
EDIT: My primary C7 got to 10 Load Alerts and rebooted - it's looking healthier now:
1 Like
Like you I have over 300 devices. I have found that the response time of some of my devices slows considerable when free memory gets around 200 MB so I reboot at 225.
1 Like
It seems like you are starting off in a hole as far as memory goes. After a reboot, my hubs are both over 400. After a couple days they might settle out around 300. If they drop too much below, then cpu usage starts climbing, I assume due to the hub trying to free up more memory. Your database is also much larger.
If I reboot a second time, I get more memory free - I have no idea why that is.
Havent noticed that. I pole once an hour both zwave and zigbee device.and the only slowdowns.are.around the 2 am maint cycle.
