Self doubt

I spent a couple hours yesterday doing some reorganization and rewriting of code. I woke up this morning to one of the new automations not working. I spent 10 minutes troubleshooting the code I thought I had messed up until I figured out a battery had died on as sensor...
Thought someone would get a good laugh out of this like I did.


It's amazing that we allude to the most complex solutions when something goes wrong. All the while the real answer was waiving it's arms and screaming "Pick me! Pick me!"
Been there, done that. Not gonna learn from it.


vis Occam's Razor :wink:

1 Like

My first stop whenever something doesn't seem to be working... We've ALL been there :slight_smile:

Nice, where did you get that battery snapshot? Although battery reporting in general is not very reliable, cuz , lithium

hey Dragon....maybe a dumb question but where/how did you get that collated view of your device's battery state?

There are a few different ways you can get data charts like that... I'm using the event socket with nodered to gather and export the data to an influxdb and then using grafana to render my data....

There are a few threads on here that can walk you through getting your own logging and charts set up in one of various ways... theres even a new app Hubigraph that lets you get some charts and graphs into HE itself

I prefer the influx->grafana route because i can have as much historical data as i want, and i can combine it with data from other sources as well.

This is just some of the data from my automation's, but i have other sets that also include data on loads and runtimes etc from my pc, 3d priners, phones, routers, pi, etc.

Edit: For those wondering because of the graphs.... I currently have 14 HEM outlets installed on my hub (yep just 1 hub). Only time lag occurs is during the maintenance window, and as long as I do a daily reboot my response time (how long it takes the app page to load) stays under 1.5 seconds, with the average being .67 seconds... If i go 4 or 5 days without a reboot my response times do rise up to 3 seconds with a 1.8s average. not the 14 second spike earlier, which was due to a manual database backup before a reboot as I noticed my average times creeping towards the 3s mark


I mean yeah its not dead on accurate but Its close enough.... As long as they are over 20% they are good. And if they are dead then its simple to tell as they will wont report, which returns a null entry, which i have set to equal 0 on the battery chart. :wink:

1 Like

woah...and here i thought you'd say "just click Battery Report hidden at the bottom a page in HE, dumbass"


you've gone all First Order on your reporting, I am impressed sir! (but i don't think i'll follow in your footsteps, too much time, effort & server building for my liking)

There are much simpler ways to get some basic data charts...

Check out hubigraph... It should do what you want without any external server or programs etc, just a couple on hub apps, :slight_smile: