Updated C-4 hub to 2.5.1.174 and received "Hub issue 2.41 Severe CPU Load detected" from a Rule Machine that monitors Location event: severeLoad and more. The really odd thing is that the Hub Info app showed the cpupct to be over 155% when this occurred.
Did it go away, Arn?
I often end up with the high load message after a reboot. When I look it doesn't look like everything is being reset to zero and the devices/apps causing the load actually have a higher accrued active time than the hub has been up. I ignore it and it goes away in about 10 minutes or so. This to me looks like the reason we see load values over 100%.
There is always the possibility I have no idea what I'm taking about but the load message always goes away and does not come back, unless I reboot.
Hello Steve
It seems to occur one time about 5 to 10 minutes after a reboot and I've also seen this a few times in the recent past. What's weird is that 155% cpu number in the HubInfo app.
I agree. The more I have loaded up my C8 the more I have seen this at reboots. It goes away within ten minutes. The UI loading the main page really contributes to it, so after a reboot, if the UI is spinning trying to load the page, I close the browser so it doesn't keep fighting to load, and the hub settles down quicker. I check back in ten minutes and everything is fine.
Linux handles multi-core processors by default as "100% per core". This means a 4-core CPU can show up to 400% usage.
Can you PM me your hub's id? I'll check the engineering logs.
PM sent since I just experienced this with the new update.
PM Sent
Well that explains the CPU % being greater than 100, and hopefully @gopher.ny will figure out why it skyrockets around 10 minutes after a reboot.
Grab 2.5.1.181, it should deal with the exceptions reported in the engineering log.
High CPU following a reboot is pretty common. It can be dependent on drivers/apps/ect that update data at reboot time. I know i have spent time tweaking my Govee code to minimize it impact during this window.
10 min after is a little odd unless you have some stuff that intentionally delays functions after reboot. I have done some of that too.
I think the Hubinfo driver also calculates its cpu% off the load value, so technically it can even go above 400%. CPU Load and CPU% are similar, but not 100% relatable as a system with high waits can well exceed the load of 1 per thread.
I'm the same. Always get high load after a reboot ever since I've had HE. (I started with a C3!!).
Not a big deal as far as I'm concerned as everything settles down after 10-15 mins and during that 15min window, things run OK anyway.
Just my experience. 
What I would be concerned about is if the load suddenly jumps for no reason when the hub is running normally after, say, an hour.
Not sure if this means anything but here goes.
On my hub the issue reports about 8minues and change after reboot.
- Reboot 21:28.46 Hubload msg 21:36:54 Time from reboot 8:08
- Reboot 11:20:17 Hubload msg 11:28:34 Time from reboot 8:17
@gopher.ny Does the hub do something at this point or perhaps a few minutes earlier after a reboot? Also when I look at scheduled tasks prior to the reboot cpuload msg it shows zero tasks. Now it shows 54! Hmmm....are all apps fired off at the same time after reboot or is there a slight delay between app starts?
My concern is that the hub might not respond to a critical condition while it is slowed down during reboot. I totally understand that things may not work as expected until all initialization completes.
I see this as well when my hubs are rebooted. It has been occuring for years. It has to to do with all of the extra stuff on reboot. How much the impact is and how much the impact is depends on a few things. Generally speaking this is a known thing on reboot. I believe that there is a built in delay after a reboot before this warning message appears.
When your hub reboots it has to re initialize all of the devices and apps. Some times this means getting most recent status, loading values, importing data from flat files. There are sevearl other things as well. I don't think the hub try's to limit how much of those tasks it does. It just lets them all go and finish as they can.
My Govee drivers have allot that needs to be loaded or initialized on reboot. Individually it doesn't seem bad, but as your device, integration, rules, app counts goes up it becomes more impactful. My guess is that it is busier before that message, it just doesn't report it until that time.
I haven't noticed this, but honestly i haven't been focused on if anything is missed while the hub is initializing the drivers. I would think it might be delayed slightly if anything.
I would expect that most of what causes the high cpu load on reboot is actually wait states as the devices reload their current states or load information from disk/remote sources. like a API call.
Have you seen any occasions where your concern has occcured.