I’ve been running the yellow for about a week now and I’m not impressed. i was previously running on a docker on my home nas, using mosquito and zigbee2mqtt to run everything else. the docker setup was flawless.
The yellow on the other hand ranges for the same performance speed wise to up to 10 seconds latency in performing automations, browsing the interface or just responding. I’m about to sell this and go back. can anyone help?
Please post a callgrind file from the profiler integration using the start service
Compare MCU on your nas and MCU on Yellow … if MCU on NAS is better/faster (which i think it is) then you have your culprit. If so, you can make wonders with HA and you won’t make Pi run faster.
I think that running HA on NAS, since you have it, is best option anyway → you have your data pretty much secured (in case of HDD failure) if nas is running any kind of raid (except RAID0, of course).
Yellow is Pi4, if i’m not mistaken. I did run HA on my Pi4 (just for test) and it ran slower than my HA inside VM on Synology.
do you have any instructions for that?
Add the integration from Profiler - Home Assistant and then call the profiler.start service. Download the files it generates listed in the notification
Disable the radarr integration. Its using a slow json and date parser which is eating up 30-40% of your cpu time. You should probably open a github issue as well with your profile file asking the author to switch to using orjson to parse json and ciso8661 to parse dates where possible
Just found it myself not too long ago, thanks so much!
@bdokka or anyone else I am having simalar issues with my Yellow, super slow not sure what is going on,
tried to do what you did here but no luck. I did get profiler loaded and I have the files. would you mind looking at /converting them for me? Thanks.
If you post a link to the files or PM the to me I can take a look.
Thanks. I “think” I may have fixed the problem but if it occurs again I will post the files. Thank you for the reply!
How did you fix the problem?
