Is it all worth it?

1

Hey everyone!

This is a bit of an existencial crisis kind of post more than a technical difficulty one. Hope you can help however you can.

So, I started exploring HA on my own in Dec 2021. Got myself a Pi 4 which is connected with a cable to my router. At that time I was especially interested in having my current Tuya/Smartlife devices working within HA. Mainly wondered whether it would be possible to have motion sensors triggering lights a bit faster, or just connect devices of different manufactures in a single automation process. I managed to configure 1 light from Tuya in HA. That was sort of my proof of concept.

Then life happened and I didn’t have the time (or will) to dedicate to HA for an entire year. The whole Local Tuya process is a bit of a cumbersome one, and not exactly bulletproof, which didn’t help either.

Either way, recently, in Dec 2022 I got plenty of time at hands and decided to take my HA to the next step. Updated HA itself, along with every extension that needed an update. Didn’t really had anything going on so the risk of having to format it and start from scratch was worth it.

By January I had configure all but 1 out of +20 Tuya devices in my HA — the 1 that didn’t was refusing to connect due to some “missing datapoints” of some sort. Tried power cycling it. Triple checked the IP was correct. Tried multiple protocols. Tried moving the router closer to it. Nothing seemed to work. The device was working fine through the Smartlife app though. And Google Home too.

Also managed to set up my Eufy cameras and homebase, making it possible to finally connect Tuya with Eufy in an automation that would flash some indoor lights after dark, when motion is sensed outside. This worked for a single day only.
Even managed to set up an automation for both my and my wife’s phones so that the Eufy homebase would disarm the alarm when both of us were at home. Which also only worked for a day.

The day after, all of a sudden, HA became wildly unresponsive. Reboots would take several minutes instead of roughly 1 minute. HACS rarely completely loaded — the icon on the sidebar would usually be hidden next to the label.
The Eufy integration stopped working.
The Logs mentioned “rate limit” with GitHub. I searched a lot and read a lot of threads in here. Including one where someone mentioned GH had an API endpoint where we would check how many requests we had left. It was all absolutely fine. Not only that, even hours and days after these log messages first pop up, it would still be unresponsive.

Did plenty of reboots. They all took several minutes.
Hassio alone was taking +3 minutes to load, with log messages such as:

Waiting on integrations to complete setup: hassio

At some point uninstalled the Eufy integration and it didn’t help. When I tried to reinstall it, got no luck as it wouldn’t go past the loading screen — presumably because of the rate limiting again.
Which means I can’t update anything right now.

To me, it all felt like this happened all of a sudden. But I may have tweaked or installed something that I can’t remember and could have been messing up something, it would only be fair to thing so.

This morning, without any progress, I decided to restore a backup 2 weeks old. I’m absolutely certain that 2 weeks ago it was working fine, even though it didn’t have some automations/integrations configured — such as the Eufy one.

It’s behaving in a very similar way. Hassio still takes a long time to run, but it’s slightly faster now, due to not all integrations being installed, I suppose. But it’s still struggling with rate limits, and that’s likely what’s causing all the delays.

Then I tried an yet earlier backup. Can’t tell any difference either. It’s so far back in time at this point that I would have to reconfigure everything I had previously done. All Tuya devices/automations. All Eufy devices/automations. Remote access. Our phones configured, etc. Might as well start from scratch!

—

If you’re still here — thank you! — the next bit is where you come in.

At this point I’m wondering if all this is worth it.
How has been your experience both while setting up as well as after it?
Do you usually install system and integration updates as they come out? Do you fix all versions and consider it “done” for life?
Do you have issues when installing new versions?
If you don’t update, do you have issues that come all of a sudden?

I can see HA being extremely useful, don’t get me wrong! But I’m not sure I want to be eternally tinkering with it just to make sure my house is properly functioning.

Thanks :pray:

2 Likes
2

Might want to take a look at this thread. In general, there is a learning curve that requires some (a lot?) of time invested. This forum does have answers to most of your questions if you search around a bit.

3

What kind of sd card are you using? Could be an sd card gone bad. Out of the box HA writes a lot to the sd cards and those things are not really build for that.

4

The very first time I set up Home Assistant on a pi3B I crashed it so bad within a week I had to start from scratch. This was 4 years ago though and things have improved considerably since then. Never did work out what went wrong.

I spent years afterwards reading documentation and programming the back and frontend to how I liked it. I made some poor hardware choices (ir control with no state feedback is bad), and learnt from my mistakes.

I’m at the point now where I have to do very little maintenance and all my hardware works wonderfully.

Yes I have alerts for updates. Once a week it takes less than five minutes to apply the updates. Ten minutes max if there are core or integration breaking changes.

Hardly ever. When I have time I participate in beta testing. If something goes wrong I roll back a version in less than 2 minutes. Less than 10 minutes if I have to restore a backup.

I did not like the recent frontend colour changes but managed to work around them.

Things I have planned to make my system even more reliable and user friendly:

Move away from wifi smart lights and install zigbee dimmer switches instead.
Simplify my dashboard themes and reliance on custom cards.

In short: it can take quite some time to get everything running smoothly, it requires careful hardware selection (no cloud based or devices without state feedback) and a lot of trial and error. But yes it is worth it. It has also been and continues to be a mostly enjoyable (sometimes challenging/frustrating) hobby.

1 Like
5

My 2c to this:

Yes, I have been on various automation platforms since 2016 and when I gave myself enough time to dig in to HA a couple year ago it’s been well worth the time. Well set-up with wise hardware decisions it’s been rock solid. But you still need to invest time in it but I have not stumbled on anything that gives you near the same capabilities that HA does.

From the beginning it was confusing, but I invested some time, as said above, and when you understand the basic concepts of HA (which for me took some time) it’s fairly strait forward and the problems I have had afterwards are some few breaking changes or my own fault.

I keep my production system updated but today never with the first update of the month e.g. 2023.1.0. I normally wait for .2 or .3 and read the forums if other users have experience any problems that could be related to my set-up. I also read the breaking changes of each release.

Rarely, no. I sometimes have issues with some integrations via HACS but they seem to be fixed fairly soon. I have had one more serious crash during two year when I was too quick to install an early .0 monthly release but when I rolled back everything was in order and the fix was issued later that month.

Normally no, but the question is not really applicable to me.

Edit, to expand on my answer:

Today, I also use a RPi4 with an external SSD for my production system and the experience for that is above. Today I have ordered a NUC to be more future proof.

6

my experience with HomeAssistant and to what it became within two years:

I was already interested in SmartHome for several years, but back then, smarthome was a VERY expensive technology, and not really something you could implement in an existing house hold.
At that time, people started with creating scripts, writing their own software for wall-mounted tablets and started developing hardware hacks, etc.

during that time, I had to create something to controll some terrariums (heating, lighting, etc.) - and decided to do it with a Siemens LOGO… as I thought, it could be more reliable than any custom built solution.

OK, I think - back then - at some point I found a Software called “HomeAssistant” - and decided to have a look into it.
Oh dear. I don’t really remember about the installation process… and I am not sure if this was one of my first “Linux experiences” since I had to install it on a VM for testing HomeAssistant.

Unfortunately, it was a frustrating experience - because: I couldn’t do anything.
I haven’t had any devices that I could use with HomeAssistant - but also, it was a mess to configure a UI… / Dashboard… AND I didn’t really had the mood to dig into it more deeply - because I just couldn’t see the benefit from it.

MANY YEARS LATER - I Moved into another house with my at that time - girlfriend.
I’ve started to “extend” my IT stuff… and over the years, I also started adding the first smarthome components into our household - starting with “Homematic IP” - which was running on a Cloud.
I am not sure, if I installed HomeAssistant on my existing Synology Disk Station back then… because it was also the time, when I came to the conclusion, that the diskstation does not fit my needs anymore…

A New system was needed - for managing my movie collection with I had built up over the last months.
It became proxmos - mainly running as a Network storage - but also running Emby on it.

Also, I’ve installed a HomeAssistant version - to connect my Homematic Stuff somehow… but since it was a Core Installation and still with nearly no experience, it was just unstable and I didn’t really used it… (also, I just had some smoke detectors, a single Light Switch and some Heat valves and Windows Sensors that were able to communicate directly)…

so… even at that time, HomeAssistant was not really something I’ve really used.
This was 2018 / 2019…

In 2020 - I’ve started to plan a Solar System on our house… And I found out, that it was possible to integrate the Inverter into this peace of Software - which was running somewhere on my network but wasn’t really usefull…

This was, when I really started with HomeAssistant - and it became a major part in my HomeNetwork.

Still running on the Proxmo Server, I did a complete new installation.
this time - running as a VM with HomeAssistant OS - more reliable, but still… far from being perfect.
In 2021, finally our Solar System has been installed - and I started to create the HomeAssistant as a Monitoring platform. The introduction of the Energy Dashboard therefore - came just at the right time :wink:
2021 was a year, where many of the integrations I had to use (for example the solar edge modbus, etc.) became stable enought to work with the energy dashboard in a reliable way) - so I would call 2021 the year of “getting stable”… 2022 was running smoth… and was he year where I extended the smarthome with more devices, optimized monitoring, got a new EV Car, a Charging station that could be implemented into HomeAssistant and got a new, smartcontrolled coffee machine.

In the last half of 2022 - I’ve decided to do a complete rework of HomeAssistant - cleaning up integrations and dashboards - since I’ve tried many things the last year - so the goal was to start 2023 with a complete, fully tested and working HomeAssistant… which becomes even more important since I’ve also started to combine Voice Control with HomeAssistant … (so again, perfect time to introduce the year of the voice) :wink: :wink:

BUT: Conclusion:
Yes, it takes time.
Yes, it is required to like the work on such things…
If I would still have devices mostly from one specific vendor, I would probably stick to their app and not care about something like HomeAssistant.
AND: I am not sure if I would use an Raspberry for a project Like HomeAssistant in that situation… the fact that I already HAD my server for other tasks - was a major issue why I started with HA…

Is it worth it?
2021 - starting to learn HomeAssistant
2022 - starting to extend the setup
2023 - ??? (having a productive setp and probably a dev environment for testing / playing etc)
Yes - it’s worth in my case.

7

yes, absolutely.

I set it up a LONG time ago. My experience with setup will be completely different than yours.

But since then I really haven’t had any major issues aside from occasionally needing to fix, or trying to find work arounds for, all of the non-stop breaking changes. Luckily most of them don’t always apply to everyone and I haven’t really been hit too hard with most of them.

I install a core update only once a month and only after it has “matured” a bit. After that update, as long as my HAworks with that version I see no need at all to update to the later monthly point releases.

As long as you don’t add any new functionality (hardware etc) to your house and any cloud based hardware doesn’t get the API changed then HA will continue to run as long as the HA hardware itself doesn’t fail. You will never need to update ever again at that point.

But if you ever do update anything outside of HA and HA doesn’t support it on the version you are running then it’s going to be WAY more difficult to update to the current version and fix everything in between.

So just update once a month and spend the few minutes to fix the breaking changes (if there are any) and be done with it. It’s way less painful that way.

And as far as tinkering unless you totally lack any imagination there’s ALWAYS going to things you will think of that would be “really cool” to add to HA or something that will just make your life easier.

Tinkering is pretty much THE point of HA.

8

Almost an entire year later… Feels like I’m on my annual internet visit :sweat_smile:

Thanks a lot for all your inputs, very much appreciated!

As I initially mentioned, I’m extremely frustrated with HA — still am, after all this time.
Meanwhile, I haven’t touched HA at all and have solely been living with Smartlife automations. Sometimes even those don’t quite work as expected or take quite a long time to respond.

At this point we’re struggling to even be alive with our newborn at hands, let alone find time to tinker with HA…
I got fed up with all this “automations” world — for now, at least — and decided I would go back instead of forward. I’m installing old school PIR sensors on a bunch of lights which happen to be installed linearly so that 1 sensor triggers 3 lights at a time.

No internet reliance — which have failed me in the past — and something that will most likely stay in house even if we move out, as no accounts or services are connected to it.

Obviously I loose the ability to trigger more than 1 group of lights when 1 sensor finds movement, as well as the ability to switch cameras off when we arrive home, but oh well… I can live with that if everything else gets fixed.

Again, thanks a lot for all your input, who knows if I’ll be back here in a year or two.
Until then :v:

1 Like
9

Yeah that is definitely going to restrict any hobby!

See you again in a few more years. :slight_smile: