# CPU load **URL:** \ **Category:** Questions & Help\ **Created:** [January 12, 2024, 8:49am UTC](https://community.homey.app/t/cpu-load/100882 "2024-01-12T08:49:08Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![Takie](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/takie/32/86573_2.png) [@Takie](https://community.homey.app/u/Takie)\ **Post date:** [January 12, 2024, 8:49am UTC](https://community.homey.app/t/cpu-load/100882/1 "2024-01-12T08:49:08Z")
My CPU load graph looks like this: ![Screenshot 2024-01-12 094000](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/8/d/8db828c4fa21bf8b97663e397a749d27a67adfe3.png) I didn’t find any suspicious individual apps though. All have low single-digit cpu loads. And when I would sum them up it would still be a far reach from the peaks I’m seeing in my overall CPU load. Is this normal behavior? If not, how can I figure out what is causing this load? Also the developer portal isn’t off much use. “App performance” shows flatline 0% all the time ---
**Author:** ![Sharkys](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/sharkys/32/54029_2.png) [@Sharkys](https://community.homey.app/u/Sharkys)\ **Post date:** [January 12, 2024, 10:36am UTC](https://community.homey.app/t/cpu-load/100882/2 "2024-01-12T10:36:14Z")
> [@Takie](#): > > Is this normal behavior? It’s not, Athom is aware but not really addressing or explaining, what would be the root cause… > [@HP23 high average CPU utilisation?](https://community.homey.app/t/hp23-high-average-cpu-utilisation/91288): > > I noticed that when HP23 has been running for a few days, the average CPU utilisation creeps up to 35% in idle mode, with periodic spikes up to 85% which seems quite high for the new hardware. When I reboot, this drops to a much more reasonable 5-10% idle with spikes up to 60% and it stays like that for a while (see screenshot). So something seems to be consuming 25% CPU continuously. Reboot at yellow line: I have about 60 Zigbee devices, mainly Aqara sensors, generic bulbs and Tuya … > [@Mysterious peaks in CPU](https://community.homey.app/t/mysterious-peaks-in-cpu/100076/6): > > Indeed recommended to create a support case with Athom. I really don’t understand why there has been absolutely no progress at all in all those months. I am still experiencing the same issues reported 2 months ago (CPU peaks every 2h + sudden high average CPU every few days). frowning > [@Please explain this pattern in System CPU load](https://community.homey.app/t/please-explain-this-pattern-in-system-cpu-load/82261/43): > > I had some correspondence with support on the increasing memory build up. There is currently no solution and they dont know what is causing this, its now with the developers and hopefully it will be improved with an update. For now the only solution support indicated is a daily restart… ---
**Author:** ![JeeHaa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/jeehaa/32/29849_2.png) [@JeeHaa](https://community.homey.app/u/JeeHaa)\ **Post date:** [January 12, 2024, 11:57am UTC](https://community.homey.app/t/cpu-load/100882/3 "2024-01-12T11:57:13Z")
@Takie Same here, 60 zigbee devices. Do you have many (Zigbee) devices on your HP23? Please also create a support case with Athom so they know that many people are affected! ---
**Author:** ![Takie](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/takie/32/86573_2.png) [@Takie](https://community.homey.app/u/Takie)\ **Post date:** [January 12, 2024, 12:17pm UTC](https://community.homey.app/t/cpu-load/100882/4 "2024-01-12T12:17:13Z")
I actually have zero direct zigbee devices connected to my HP23. Only via the Philips Hue hub. Now on wifi, will try on ethernet one of the next days. Who knows maybe that’s causing the high load. ---
**Author:** ![Peter\_Kawa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/peter_kawa/32/173848_2.png) [@Peter\_Kawa](https://community.homey.app/u/Peter_Kawa)\ **Post date:** [January 12, 2024, 1:18pm UTC](https://community.homey.app/t/cpu-load/100882/5 "2024-01-12T13:18:57Z")
Hmmm, it’s just over 50% every now and then. Not an issue imho (correct me if I’m wrong here). But I can imagine one wants to know the cause. These are my 10 year old NUC cpu’s 24/7 😬 ![Screenshot from 2024-01-12 14-16-57](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/1/5/156bccb31e1da7f4aba4f8b12ea5b2103ae303fe.png) ---
**Author:** ![Takie](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/takie/32/86573_2.png) [@Takie](https://community.homey.app/u/Takie)\ **Post date:** [January 12, 2024, 1:41pm UTC](https://community.homey.app/t/cpu-load/100882/6 "2024-01-12T13:41:56Z")
Not really clear on the screenshot, but the peaks go up into the high 70s. I’m new to Homey and only have a small amount of apps, devices and flows setup. Planning to add a lot more, so for sure this can become a big problem. Also I would very much like to know where all the load is coming from. ---
**Author:** ![Peter\_Kawa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/peter_kawa/32/173848_2.png) [@Peter\_Kawa](https://community.homey.app/u/Peter_Kawa)\ **Post date:** [January 12, 2024, 1:47pm UTC](https://community.homey.app/t/cpu-load/100882/7 "2024-01-12T13:47:12Z")
> [@Takie](#): > > Also I would very much like to know where all the load is coming from. Yeah I fully get that And true that, the insights graphs suck at displaying peaks ---
**Author:** ![Takie](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/takie/32/86573_2.png) [@Takie](https://community.homey.app/u/Takie)\ **Post date:** [January 12, 2024, 1:47pm UTC](https://community.homey.app/t/cpu-load/100882/8 "2024-01-12T13:47:58Z")
Here it is overlayed with all my apps: ![cpu2](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/2/6/26aba79454ac6468ef0c052a5f3234b70c4d5171.png) ---
**Author:** ![Sharkys](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/sharkys/32/54029_2.png) [@Sharkys](https://community.homey.app/u/Sharkys)\ **Post date:** [January 12, 2024, 1:48pm UTC](https://community.homey.app/t/cpu-load/100882/9 "2024-01-12T13:48:21Z")
I’m concerned personally about CPU peaks just because of the temperatures, as it’s connected… after all, those peaks shall not represent anything wrong but still - strange that nobody knows because then nobody might be seeing potential problem behind 🙂 ---
**Author:** ![Sharkys](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/sharkys/32/54029_2.png) [@Sharkys](https://community.homey.app/u/Sharkys)\ **Post date:** [January 12, 2024, 1:49pm UTC](https://community.homey.app/t/cpu-load/100882/10 "2024-01-12T13:49:18Z")
Yes, it’s clearly homey core… (homey-pro container, better to say) ---
**Author:** ![Peter\_Kawa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/peter_kawa/32/173848_2.png) [@Peter\_Kawa](https://community.homey.app/u/Peter_Kawa)\ **Post date:** [January 12, 2024, 2:05pm UTC](https://community.homey.app/t/cpu-load/100882/11 "2024-01-12T14:05:51Z")
Yeah it’s a bit odd the manufacturer does not know the cause 🫣 ---
**Author:** ![Glenn\_Seegers](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/glenn_seegers/32/8353_2.png) [@Glenn\_Seegers](https://community.homey.app/u/Glenn_Seegers)\ **Post date:** [January 12, 2024, 11:24pm UTC](https://community.homey.app/t/cpu-load/100882/12 "2024-01-12T23:24:32Z")
![IMG_4238](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/b/d/bd9d1cf2f10bd1ae50de73bb064b897ae35c902a.jpeg) Since the latest RC its even worse😇 before it was occasionally but now its all the time (i am in contact with athom about it btw) ---
**Author:** ![Robin\_De\_Lange](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/robin_de_lange/32/25173_2.png) [@Robin\_De\_Lange](https://community.homey.app/u/Robin_De_Lange)\ **Post date:** [January 13, 2024, 6:56am UTC](https://community.homey.app/t/cpu-load/100882/13 "2024-01-13T06:56:02Z")
same here, like to hear what answer you get from Athom. ![Schermafbeelding 2024-01-13 om 07.55.00](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/3/4/348356ff5c9c7b02e468a1737be13899e5ca0e81.png) ---
**Author:** ![JeeHaa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/jeehaa/32/29849_2.png) [@JeeHaa](https://community.homey.app/u/JeeHaa)\ **Post date:** [February 8, 2024, 10:50am UTC](https://community.homey.app/t/cpu-load/100882/14 "2024-02-08T10:50:25Z")
FYI: > [@Mysterious peaks in CPU](https://community.homey.app/t/mysterious-peaks-in-cpu/100076/7): > > Looks a lot better since I installed the latest v10.3.0-rc.12 update yesterday: [image] v10.3.0-rc.12 Core Adds optimisations that reduce the amount of data written to persistent storage. Updates translations. Zigbee Fixes an issue where Zigbee would stop working when Thread started. Not sure how this issue went unnoticed by Athom for so long, if it was caused by excessive writes? It was already mentioned here: [https://community.homey.app/t/modding-howto-investigating-homey-s-emmc-wear-e…](https://community.homey.app/t/modding-howto-investigating-homey-s-emmc-wear-estimation/101149/20) ---
**Author:** ![Takie](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/takie/32/86573_2.png) [@Takie](https://community.homey.app/u/Takie)\ **Post date:** [February 11, 2024, 1:38pm UTC](https://community.homey.app/t/cpu-load/100882/15 "2024-02-11T13:38:52Z")
Bij mij op v10.2.1 lijkt het nu ook prima te zijn: ![image](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/1/8/1838ec5582290f8a3c5ebbebf573cdb9abf997d6.png) ---
**Author:** ![Arjan\_Neet](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/arjan_neet/32/17808_2.png) [@Arjan\_Neet](https://community.homey.app/u/Arjan_Neet)\ **Post date:** [October 17, 2025, 1:54pm UTC](https://community.homey.app/t/cpu-load/100882/16 "2025-10-17T13:54:38Z")
Where exactly can you see this? I really have a lot of issues since a few weeks (mainly huge delays) and can’t find the issue, even got a loaner from Athom, but if I restore my backup on that the same thing happens but more frequently (say all the time) while with my own homey it happens a few times a day. I notice it immediately because all my Niko hardware starts blinking red, and well also the lights don’t go on anymore when I walk in a room. If I look on the homey webpage on that moment I have seen cpu average over the last 5 minutes going up to even 117% (don’t know how that is possible), but mainly around 80 to 90% while normally it is more around 20 to 30%. Have already disabled various apps and flows, but issue seems to be just everything zigbee, thread and wifi works fine (govee for example) ![image](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/8/7/871098ae4650c66160b0bfe64817d609262d8cbd.png) ![image](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/2/e/2e479b052b13bd7f374411639358e5d1fe1a1678.png) ---
**Author:** ![Peter\_Kawa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/peter_kawa/32/173848_2.png) [@Peter\_Kawa](https://community.homey.app/u/Peter_Kawa)\ **Post date:** [October 17, 2025, 2:16pm UTC](https://community.homey.app/t/cpu-load/100882/17 "2025-10-17T14:16:50Z")
> [@Arjan\_Neet](#): > > Where exactly can you see this? [https://my.homey.app/insights](https://my.homey.app/insights) . > [@Arjan\_Neet](#): > > issue seems to be just everything zigbee, thread and wifi works fine. When this means: all zigbee stuff slow things down, but Thread doesn’t: This is because zigbee and Thread share the same chip, called _ **MultiProtocol** _. In general it’s not recommended to use both at the same time, _ **only Athom** _ thinks it’s no problem and doesn’t mention the restrictions. This is what the chip manufacturer shares about it: ![Screenshot_20251017_162146_Chrome](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/e/b/eba0abc5da7428007695bd308a2b4b52b3517554.jpeg) Try to disable Thread, to see if zigbee starts behaving as should. Starting from zero might be a plan as well; To try to find the cause, you can start by disabling _all_ apps, and then enable them one by one, with maybe 10 - 15 minutes in between, and check the system insights graph, or cpu . Other system “killers” I found out, too many triggerd flows by “the Power changed” (this triggers _ **every sec** _), caused the system load to rise. Reaching 300% was no problem (Pro 2019). ---
**Author:** ![Arjan\_Neet](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/arjan_neet/32/17808_2.png) [@Arjan\_Neet](https://community.homey.app/u/Arjan_Neet)\ **Post date:** [October 18, 2025, 10:00am UTC](https://community.homey.app/t/cpu-load/100882/18 "2025-10-18T10:00:08Z")
thanks, I never realized that I could also see system diagnostics in there, thought it was only for energy stuff etc 😖 That said, if I look at the CPU usage for the past few days it says maximum 40%, so how would it then be possible that I see over 100% in the average values directly on the system settings?? I definitely do not have an every second trigger, most are just triggered by motion or by certain time. I did disable ALL apps BUT niko and hue without a bridge, since those are actually (almost) the only zigbee devives that I use, the rest is mainly wifi stuff, but that didn’t help either. At the moment since I saw so many problems with the niko stuff, I disabled the niko app and restarted homey, and so far I have not noticed anymore delay, of course it is only a short time, so it is early to say, but the last few days I had the issues like 10 to 15 times a day. If there are no more issues till tonight I will try to enable the Niko app again to see if the issues return again. I was finally happy that the non working niko buttons issue was solved in V1.7.0, but apparently a new one appeared….see that it is now at v2.0.2 Regarding thread, I only have 2 zemismart roller shutter motors to open and close the blinds, but that works only half of the time since they disconnect all the time (and reconnect after a while too), Athom advised to check the settings of my border router, I thought homey is the border router? I’m not that familiar yet with thread/matter. Anyway I disabled that flow already and just open them by remote now. ---
**Author:** ![Peter\_Kawa](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/peter_kawa/32/173848_2.png) [@Peter\_Kawa](https://community.homey.app/u/Peter_Kawa)\ **Post date:** [October 18, 2025, 12:48pm UTC](https://community.homey.app/t/cpu-load/100882/19 "2025-10-18T12:48:47Z")
> [@Arjan\_Neet](#): > > if I look at the CPU usage for the past few days it says maximum 40%, so how would it then be possible that I see over 100% in the average values directly on the system settings?? Athom mixes the descriptions “CPU usage” and “System load” or “Average load” → CPU usage indeed cannot exceed 100% → System load can reach 300% or more Proof Mobile app: ![Screenshot_20251018_142028_Homey](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/3/a/3ae708a27336a4b994c9267a4f8ab75068f0c05f.jpeg) Web app: ![Screenshot_20251018_141940_Firefox](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/c/a/ca640e852e3da6bd947608ab5595555c79dd534b.jpeg) My Pro 2019 does not have CPU figures, but it does have this “average load”; however, this is very different from the above % values: ![Screenshot_20251018_143135_Firefox](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/9/5/95ca39596c44c866390d2088aab3a43a34ec6289.jpeg) There is also a handy monitor app available: [SysInternals](https://community.homey.app/t/app-pro-sysinternals-shows-homey-s-system-internals/78650) . > [@Arjan\_Neet](#): > > many problems with the niko stuff A possible cause can be “chatty” devices, same as can occur with z-wave. Certain Tuya devices are known to ruin / clutter the zigbee mesh by emitting way to much messages; probably certain Niko device types do this as well. I’ve also read about a dimmer doing this. Homey just tries to keep up, wich can cause these higher system loads. BTW a CPU running at 100% is nothing weird, it only throttles down when there’s less to process, to save energy . . > [@Arjan\_Neet](#): > > Athom advised to check the settings of my border router, I thought homey is the border router? I’m not that familiar yet with thread/matter. Depends on how you paired them. But the description sounds as a range problem, the roller shutters seem to be on the edge of the range. And the range can always vary a bit, due to interference. ---
**Author:** ![Arjan\_Neet](https://sea1.discourse-cdn.com/flex025/user_avatar/community.homey.app/arjan_neet/32/17808_2.png) [@Arjan\_Neet](https://community.homey.app/u/Arjan_Neet)\ **Post date:** [October 18, 2025, 1:28pm UTC](https://community.homey.app/t/cpu-load/100882/20 "2025-10-18T13:28:48Z")
the 2 roller shutters are actually only 3 meters from my homey in clear sight… The sysinternals app I’m going to give a try, seems interesting edit: actually I only see 2 cards which don’t seem that helpful? ![image](https://us1.discourse-cdn.com/flex025/uploads/athom/original/3X/1/8/18f4102e914c73508146c8793d9cada210243a48.png) The only tuya devices I have at the moment are 2 contact switches and 2 buttons, but the issue I had was also when the tuya app was disabled. Still till now I have had no issues so far just by disabling the Niko app… [Next page](https://community.homey.app/t/cpu-load/100882.md?page=2)