Hi Guys.
Currently, my Raspberry Pi 4 is only running Home Assistant OS, which works quite well.
In the last few weeks, I have grown more and more interested in paperless ngx and Nextcloud. Do you think my pi would be able to handle home automation and some light document management simultaneously?
If so, should I install a fresh version of Raspbian or let HA-OS handle the Docker containers?
Thanks for your answers, and have a wonderful day :)
Ive been running HA (in docker container) and NextCloud on rpi4 4gb and it was working fine, but after server upgrade (diy pc) nextcloud became much more snappy (wasnt my reason to upgrade). Not sure about paperless, but I had many more lightweight services along with HA and NC.
Im also not heavy HA or NC user, just my 0.02
I would, but you have to decide on your own. I dont know can you afford HA downtime, how much experience you have and how much time you want to burn. I found docker frustrating to learn but super rewarding. Keep in mind Im still new.
Its best to backup and/or get spare SD card to learn how to setup everything. Be prepared for begginer mistakes, but once you get familiar a bit it will be much easier.
If you can manage docker containers in HAOS just try it. I have 0 experience with HAOS so cant help much, but I saw Portainer is available as HA addon. You should be able to manage all docker containers with simple Portainer GUI.
I had dietpi or raspbian OS with docker, docker-compose and Portainer installed. Then I started many containers (including HA) using Portainer. If you go this path you will lose HA addons. I believe most addons can be replaced with docker containers from hub.docker.com, but Im not sure about that. Ive just seen some HAOS vs Docker posts
Same here. My old Pi 4, was running HA, Nextcloud, Ghost blog, websites, Jellyfin and a couple of other small things. It did it fine, but when I upgraded to the Odroid H3 it all became so much faster and snappier.
I run nextcloud with two 6TB drives on a core2 quad optiplex with 4gb of ram. Runs like a dream.
I say start on the pi and migrate to something faster if you really find you want/need nextcloud.
It’s probably going to be ok to scope them out and experiment a bit, but I doubt you’ll get enough performance and stability to run it as production. Paperless’ OCR is quite heavy on the CPU - iirc you can disable it but then you lose half of what makes it useful, and Nextcloud also does some processing to files that are uploaded to it. Since you are not running pi-hole or other latency-sensitive services it will probably be fine, it will just get sluggish while it processes uploads.
While paperless processing is indeed quite intensive, it’s not like this is a latency-sensitive task. If it takes 5m to OCR a scan, so be it. That doesn’t make it unusably slow.
What I meant is that overloading the CPU on a Raspberry running pi-hole will make the whole network misbehave and timeout, until DNS requests are able to be serviced again. But since they’re not doing that it should be fine :)
I would say not in a way that makes sense, there may be hacky workarounds like setting nice priorities or messing around with scheduling, but there’s no way around hardware limitations. The Pi’s CPU, RAM, and IO bandwidth are what they are, and you need overhead to guarantee “snappiness”
Sure it is… especially when you start to have tons of JavaScript erros in your browser console :D Now seriously, have a look at this: https://lemmy.world/comment/346174
Eh, every page has a bunch of Javascript errors and logs. That’s normal. So far it hasn’t impacted my usage at all and I’ve got 450GB gigabytes stored in a Nextcloud instance that’s lasted 4-years and migrated between 3 separate machines.
Ahaha. The current state of software development and what people accept as normal is just mind-boggling. And no, it’s not “normal” nor “ok”, it slows down things and the UI sometimes crashes with more data.
I take it you’ve never seen the amount of crap that gets put into journalctl then. I’ve never encountered a Nextcloud crash with the amount of data I have. It could have been a deployment issue on your end. My docker containers have been running fine.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !selfhosted@lemmy.world
A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don’t control.
Rules:
Be civil: we’re here to support and learn from one another. Insults won’t be tolerated. Flame wars are frowned upon.
No spam posting.
Posts have to be centered around self-hosting. There are other communities for discussing hardware or home computing. If it’s not obvious why your post topic revolves around selfhosting, please include details to make it clear.
Don’t duplicate the full text of your blog or github here. Just post the link for folks to click.
Submission headline should match the article title (don’t cherry-pick information from the title to fit your agenda).
No trolling.
Resources:
selfh.st Newsletter and index of selfhosted software and apps
Acronyms, initialisms, abbreviations, contractions, and other phrases which expand to something larger, that I’ve seen in this thread:
4 acronyms in this thread; the most compressed thread commented on today has 11 acronyms.
[Thread #94 for this sub, first seen 31st Aug 2023, 10:15] [FAQ] [Full list] [Contact] [Source code]
Ive been running HA (in docker container) and NextCloud on rpi4 4gb and it was working fine, but after server upgrade (diy pc) nextcloud became much more snappy (wasnt my reason to upgrade). Not sure about paperless, but I had many more lightweight services along with HA and NC. Im also not heavy HA or NC user, just my 0.02
So I should look into hosting docket on my pi and upgrade as soon as I feel like it?
I would, but you have to decide on your own. I dont know can you afford HA downtime, how much experience you have and how much time you want to burn. I found docker frustrating to learn but super rewarding. Keep in mind Im still new.
Its best to backup and/or get spare SD card to learn how to setup everything. Be prepared for begginer mistakes, but once you get familiar a bit it will be much easier.
If you can manage docker containers in HAOS just try it. I have 0 experience with HAOS so cant help much, but I saw Portainer is available as HA addon. You should be able to manage all docker containers with simple Portainer GUI.
I had dietpi or raspbian OS with docker, docker-compose and Portainer installed. Then I started many containers (including HA) using Portainer. If you go this path you will lose HA addons. I believe most addons can be replaced with docker containers from hub.docker.com, but Im not sure about that. Ive just seen some HAOS vs Docker posts
Gl
Same here. My old Pi 4, was running HA, Nextcloud, Ghost blog, websites, Jellyfin and a couple of other small things. It did it fine, but when I upgraded to the Odroid H3 it all became so much faster and snappier.
Can always get a NUC or another Pi.
DIdn’t Intel stop making NUCs?
Intel did but believe they are being made (MSI et )
Yes but you can find old hardware for sale too. Another alternative are the micro form factor Dell Optiplex PCs.
I run nextcloud with two 6TB drives on a core2 quad optiplex with 4gb of ram. Runs like a dream. I say start on the pi and migrate to something faster if you really find you want/need nextcloud.
They did, which is a huge shame. You can get equivalent mini PCs from vendors like Minisforum tho.
It’s probably going to be ok to scope them out and experiment a bit, but I doubt you’ll get enough performance and stability to run it as production. Paperless’ OCR is quite heavy on the CPU - iirc you can disable it but then you lose half of what makes it useful, and Nextcloud also does some processing to files that are uploaded to it. Since you are not running pi-hole or other latency-sensitive services it will probably be fine, it will just get sluggish while it processes uploads.
While paperless processing is indeed quite intensive, it’s not like this is a latency-sensitive task. If it takes 5m to OCR a scan, so be it. That doesn’t make it unusably slow.
What I meant is that overloading the CPU on a Raspberry running pi-hole will make the whole network misbehave and timeout, until DNS requests are able to be serviced again. But since they’re not doing that it should be fine :)
It’s there any way around this? I don’t want my smart home applications to run sluggish. They need to have priority.
This is what the nice command is for
Scheduling priority on Linux is borderline broken. Nice doesn’t even do anything noticeable on modern systems.
I would say not in a way that makes sense, there may be hacky workarounds like setting
nicepriorities or messing around with scheduling, but there’s no way around hardware limitations. The Pi’s CPU, RAM, and IO bandwidth are what they are, and you need overhead to guarantee “snappiness”deleted by creator
Can you tell me a good scanner with integrated OCR? I used to love vflat, but they took a dark turn with a pricey subscription plan
deleted by creator
Nextcloud will be slow. To be realistic is it always slow in all hardware.
It will be slow on the Pi, but it’s pretty darned fast on my new DIY NAS with a Ryzen 5600x so it does scale with better hardware.
Sure it is… especially when you start to have tons of JavaScript erros in your browser console :D Now seriously, have a look at this: https://lemmy.world/comment/346174
Eh, every page has a bunch of Javascript errors and logs. That’s normal. So far it hasn’t impacted my usage at all and I’ve got 450GB gigabytes stored in a Nextcloud instance that’s lasted 4-years and migrated between 3 separate machines.
Ahaha. The current state of software development and what people accept as normal is just mind-boggling. And no, it’s not “normal” nor “ok”, it slows down things and the UI sometimes crashes with more data.
I take it you’ve never seen the amount of crap that gets put into
journalctlthen. I’ve never encountered a Nextcloud crash with the amount of data I have. It could have been a deployment issue on your end. My docker containers have been running fine.