Wall Mounted Dashboard (now known as HADashboard)

only:

runtimeError: no acceptor (port is in use or requires root privileges)

and then a lot of lines which give all different files like:

var/lib/gems/2.01./gems.eventmachine …

So if it ran ok the first time without root, it seems like the port is still in use :wink:

Can you find the process running thin and kill it? Alternatively you could try a reboot to clear the port.

1 Like

i rebooted.
didnt want to do that because it takes a lot off time killing everything and restarting everything :wink:
should have known that i should try that first before crying out :wink:

but i hope this will keep on going right.

my rpi is running between 70 and 95% cpu usage

Wow, that’s pretty heavy utilization. Have you thought about getting a second PI and offloading some of the workload to the second PI?

right now i am starting to think about that.

all that is running is hass and all that i need with hass so, hadashboard, appdaemon and hapush.

And that’s running 70-95% utilization? Wow. I’m just running hass, and bind on mine. Any idea which one of (hadashboard, appdaemon and hapush) are jacking up the utilziation?

with dashboard and hass it was between 35 and 45 most of the time.
i have appdaemon installed on the rpi today (before it was on my pc)
it has gone down a bit now to between 50 and 60.
but the dashboard is timing out and takes forever to load and doesnt load like it should.

Hmm. I’m running the following on my Pi:

HA
HA Dashboard
HA Push
AppDaemon (Again finally, after a long session yesterday!)
Harmony API
FTP

I’ve had a few transient peaks around 75% but they are brief and look to be restarts. I’m usually around 9% utilization. I’m surprised to see your numbers.

how can i find the proces running thin?

dashboard was hanging and i tried restarting but the same error, so the port is still in use.

dont want to restart the whole rpi again

maybe its because i have a shitload off sensors and apps running?

I have a shit ton of sensors and before the recent recode of Weather Underground, I had a huge amount of template sensors - like 15 just for that. But as you know I only got AD running last night and it’s only running the HelloWorld app. That may change once I start getting apps installed…

try:

hass@Pegasus:/export/hass/src/home-assistant.github.io/source/_ecosystem$ ps -ef | grep thin
root      3943     1  0 Nov21 ?        00:11:23 thin server (0.0.0.0:3030)
hass     32076  3821  0 10:09 pts/1    00:00:00 grep --color=auto thin
hass@Pegasus:/export/hass/src/home-assistant.github.io/source/_ecosystem$

FWIW, when I was running all of the above on my PI, my experience was similar to Rob’s.

thx andrew that made the port free.

could it be because i dont have it running as daemon?
except hass?

or is it because i get 60 to 80 sensorupdates every minute right now?

Umm, that might play a part in it… :astonished:

that happens once in a while when a temp is in the middle off 0.1 and 0.2 or 0.2 and 0.3
then the temp starts switching.

i need to restrict those updates but dont want to restrict them to much because they run my heating.

but then again even when i restrict that, i have so much sensor data coming in.
even when i restrict them all to once a minute it will still be that amount and thats only a small part off what i have in mind.

The biggest drawback of the PI platform is that it’s “disk” is just a memory card - that was one of the reasons I switched to a more powerful box. I never had any issues but I wanted room to grow. All of those updates could conceivably be maxing out your disk bandwidth and slowing everything down.

I switched to a Beebox - it’s still embedded and fanless, but it has a mobile celeron processor and an SSD.

thats is probably the problem andrew.

i have quite a few apps which are writing data to the “disk”
partly to make new sensors for hass and partly to collect data outside the big hass db.

when i take out appdaemon the cpuuse is reduced to “normal”
so i have started appdaemon on my pc again, and need to think about rewriting some apps.

but even after that, i cant get dashboard to run normal again. it times out now that i have the updated version off hass on the pi running and pointed to that. :cry:

edit: i pointed to the old hass on my pc. dashboard functions normal
poited to the hass version on the rpi again, dashboard times out.

i am afraid that the rpi isnt gonna do it anymore for me :cry:
beebox is to expensive im afraid.

Yep, looks like the poor PI is on it’s knees :frowning:

Hi can you post the instructions for the inputslider. I want to use it for my lights instead of the hadimmer.
I’ve found this one https://github.com/FlorianZ/hadashboard/pull/29/files#diff-655b9f2014d8e068661eff29667b6c13 but if you have a working solution with hapush that would be nice

sorry, but hapush doesnt give state changes for inputsliders.
hapush and the homeassistant.rb have to be edited for that.
thats why i havent given the code for this widget yet, because you would lose it after an update from hadashboard.

as soon as @aimc has a little time for it i will give him the code so that he can evaluate it and put it in hadashboard if he likes.