Without a doubt a lot do, but I personally couldn’t care less. I have a server at home, but that’s just a necessary evil. If I could I’d just rent hardware for everything, but there’s technical and obviously financial limitations with that.

And hosting pretty much anything is practically identical regardless of the platform. Sure, there’s exceptions, like my Home Assistant server with z-wave, which needs to be physically nearby my other stuff, but things like fediverse instances and other browser-based stuff are exactly the same to maintain regardless of the underlying platform.

My personal opinions, not facts:

For hdd’s to be used as long term storage, what is usually the rule of thumb? Are there any recommendations on what drives are usually better for this?

Anything with a long history, like HGST or WD (red series preferably). Backblaze among others publish their data on longevity of drives, so look for what they’re offering. On ebay (and others) there’s refurbished drives available which are pretty promising, but I have no personal experience on those.

Considering this is going to store personal documents and photos, is RAID a must in your opinion? And if so, which configuration?

Depends heavily on your backup scheme, amount of data and available bandwidth (among other things). Raid protects you against a single point of failure on storage. Without raid, you need to replace the drive, pull data back from backups and while that’s happening you don’t have access to the things you stored on the failed disk. With raid you can keep using the environment without interruptions while waiting for a day or two for a replacement. If you have fast connection which can download your backups in less than 24 hours it might be worth the money to skip raid, but if it takes a week or two to pull data back, then the additional cost of raid might be worth it. Also, if you change a lot of data during the day, it’s possible that a drive failure happens before backup is finished and in that case some data is potentially lost.

On which level of RAID you should use, it’s a balancing act. Personally I like to run things with RAID5 or 6 even if I have a pretty decent uplink. Also, you need to consider what’s the acceptable downtime for your services. If you can’t access all of your photos in 48 hours it’s not a end of the world, but if your home automation is offline it can at least increase your electric bill for some amount and maybe cause some inconvenience, depending on how your setup is built.

And in case RAID would be required, is ubuntu server good enough for this? or using something such as unraid is a must?

Ubuntu server is well enough. You can do either sofware raid or LVM for traditionald RAID setup or opt for a more modern approach like zfs.

I was thinking of probably trying to sell the 1660 super while it has some market value. However, I was never able to have the server completely headless. Is there a way to make this happen with a msi tomahawk b450? Or is only possible with an APU (such as 5600g)?

No idea. My server has a on board graphics, but I haven’t used that for years. But it’s a nice option to have in case something goes really wrong. You can still sell your 1660 and replace that with the cheapest GPU you can find from ebay/whatever, at least as long as you’re comfortable with the console you can fix things with anything that can output plain text. If your motherboard has separate remote management (generally not available in consumer grade stuff) it might be enough to skip any kind of GPU, but personally I would not have that kind of setup, even if remote management/console was available.

If you guys find any glaring issues with my setup

I don’t know about actual issues, but I have spinning hard drives a lot older than my kids which still run just fine. Spinning rust is pretty robust (at least in sub 4TB capacity), so unless you really need the speed traditional hard drives still have their place. Sure, a ton more of spinning drives has failed on me than SSD’s, but I have working hard drives older than SSD as a technology has been around (at least in the sense of what we have now), so claiming that SSD’s are more robust (at least on my experience) is just a misnderstood statistics.

this will limit ZFS ARC to 16GiB.

But if I have 32GB to start with, that’s still quite a lot and, as mentioned, my current usage pattern doesn’t really benefit from zfs over any other common filesystem.

As for using a simple fs on LVM, do you not care about data integrity?

Where you get that from? LVM has options to create raid volumes and, again as mentioned, I can mix and match those with software raid however I like. Also, single host, no matter how sophisticated filesystems and raid setups, doesn’t really matter when talking about keeping data safe, that’s what backups are for and it’s a whole another discussion.

ZFS in general is pretty memory hungry. I set up my proxmox sever with zfs pools a while ago and now I kind of regret it. ZFS in itself is very nice and has a ton of useful features, but I just don’t have the hardware nor the usage pattern to benefit from it that much on my server. I’d rather have that thing running on LVM and/or software raid to have more usable memory for my VM’s. And that’s one of the projects I’ve been planning for the server, replace zfs pools with something which suits my usage patterns better, but that’s a whole another story and requires some spare money and some spare time, which I don’t really either at hand right now.

Steps 1, 2, 4, 5 and 7 just need some time. I have the stuff pretty much thought out and it’s just a matter of actually doing the things. I was sick majority of November, but if it wasn’t for that those would have already been completed. The rest need either planning or money. Immich setup would ideally need 2x2TB ssd drives (on raid1 setup) but that’s about 500€ out of the pocket and home assistant setup needs time to actually work with it and to plan things forward. Additionally HA setup could use a floor thermostat or two, some homeESP gadgets and so on, so it needs some money as well.

Majority of the stuff should be taken care of until February, the rest is more or less open.

A ton.

1. Set up email and website hosting on a VPS to replace current setup
2. Get more solid state storage for my home server and finnish immich setup (import photos and all that)
3. Set up proper backups for the home server
4. Migrate current Unifi controller to home server
5. Local VPN server to access home assistant and other services even when travelling
6. Spend some time with my home assistant server, fine tune automations, add some more, add sensors and more controls, maybe add a wall mounted tablet for managing the thing and so on, it’ll never end and need a visit or two from electrician too
7. Better isolation for IOT things on my network. I already have separate VLAN for them without internet access, but it’s a bit incomplete project

And then “would be nice” stuff:

1. Switch Dahua NVR to something else. Current one works in a sense that it stores video, but movement tracking isn’t really perfect and the whole individual NVR box is a bit lacking both in speed and in features
2. Replace the whole home server (currently running proxmox, which in itself is fine). It’s a old server I got from work, and it does work, but it’s not reundant and it’s getting old. So something less power hungry and less noisy would be nice. It just asks some money and time, which I have neither in surplus, so we’ll see.
3. Move home assistant from a raspberry pi to the home server. Maybe add zigbee capabilities next to z-wave and wifi.

And likely a ton more which I don’t remember right now. Money and specially spare time to tinker are just lacking.

Use the friend’s network as a VPN/proxy/whatever to obscure my home IP address

And then your friend is responsible for your actions on the internet. The end goal you described is so vague that at least I wouldn’t let your raspberry connect on my network.

There’s a ton of VPN services which give you the end result you want without potential liability or other issues for your friend. If you just want to tinker, this thread has quite a bit of information to get you started.

So, you want the traffic to go other way around. Traffic from the HomeNet should go to the internet via FriendNet, right? In that case, if you want the raspberry box to act as a proxy (or vpn) server, you need to forward relevant ports on the FriendNet to your raspberry pi so that your HomeComputer can connect to the raspberry box.

Or you can set up a VPN and route traffic trough that to the other way. Tunnels work both ways, so it’s possible to set up a route/http proxy/whatever trough the VPN tunnel to the internet, even if the raspberry box is the client from VPN server point of view.

I don’t immediately see the benefit of tunneling your traffic trough the FriendNet to the internet, unless you’re trying to bypass some IP block of something other potentially malicious or at least something being on the gray area. But anyways, you need a method for your proxy client to connect to the proxy server. And in generic consumer space, that needs firewall rules and/or port forwarding (altough both are firewall rules, strictly speaking) so that your proxy server on raspberry box is visible to the internet in the first place.

Once your proxy server is visible to the internet it’s just a matter of writing up few scripts for the server box to send a message to the client end that my public IP is <a.b.c.d> and change proxy client configuration accordingly, but you still need some kind of setup for the HomeNet to receive that, likely a dynds-service and maybe some port forwarding.

Again, I personally would set up something like that with a VPN tunnel from raspberry box to the HomeServer, but as I don’t really undestand what you’re going after with setup like this it’s impossible to suggest anything else.

So, you want a box which you can connect to any network around and then use some other device to connect to your raspberry box which redirects your traffic trough your home connection to the internet?

The easiest (at least for me) would be to create VPN server on your home network. Have a dyndns setup on your home network to reach it in the first place, open/redirect a port for openvpn (or whatever you like) and have a client on raspberry running on it. After that you can connect your other device to the raspberry box (via wifi or ethernet) and create ip-forwarding/NAT rules for your traffic so that everything goes to the raspberry box, then to your home server via VPN tunnel and from there to the internet.

You can use any HTTP proxy with this, or just let the network do it’s thing and tunnel everything via your home connection, but in either case the internet would only see your encrypted VPN traffic to your home network and everything else is originated from your home connection.

You can replace VPN with just HTTP proxy, but both are pretty close the same on the terms of ‘cost’, so your network latency, bandwidth and other stuff doesn’t really change regardless of the approach. But if you just want the HTTP proxy you can forward a port on your home network for the proxy and just use that on your devices without raspberry box and achieve the very same end result without extra hardware.

And obviously, if you go with VPN tunneling for everything, you don’t need raspberry for that either, just a VPN client which connects to your home network and that’s it. The case where you have devices which can’t use VPN directly would benefit from the raspbery box, but if you already can set up a HTTP proxy for the thing you’re actually using, I don’t see the benefit of running a separate hardware for anything.

Some port forwarding or opening ports from firewall is needed on any scenario. But there’s a ton of options to limit access from anyone accessing your stuff. However, this goes way beyond the scope of your question and more details are necessary on what you’re actually trying to achieve with setup like this.

I really like the project and have been happily running it on my home lab for quite a while. But for enterprise their pricing for enterprise use is not really cheap either. 510€/socket/year is way more than the previous vmware deal we’re running. Apparently broadcom has changed their pricing to per core which is just lunatic (it would practically add up to millions per month on our environment), so it’s interesting to see what’s going to happen when our licenses expire.

As you can connect to the internet you can also access your router (or at least a router). And when running ping, even if you had overlapping IP addresses you should still get responses from the network.

So, two things come to mind: Either your laptop is running with a different netmask than other devices which causes problems or you’re connected to something else than the local network you think you are. Changes on DHCP server or misconfigured network settings on the laptop might cause the first issue. The second might be because you’re connected to your phone AP, some guest network on your devices or neighbors wifi by accident (multiple networks with same SSID around or something like that).

Other might be problems with mesh-networking (problem with ARP tables or something) which could cause issues like that. That scenario should get fixed by reconnecting to the network, but I’ve seen bugs in firmware which causes errors like this. Have you tried to restart the mesh-devices?

Is it possible that your laptop has enabled very restrictive firewall rules for whatever reason? Check that.

And then there’s of course the long route. Start by verifying that you actually have IP address you assume you have (address itself, subnet, gateway address). Then verify that you can connect to your router (open management portal, ping, ssh, all the things). Assuming you can, then check the router interface and verify that your laptop is shown there as a dhcp-client/connected device (or whatever term that software uses). Then start to ping other devices on your network and also ping your laptop from those devices and also verify that they have addresses you assume (netmask/gateway included).

And so on, one piece at the time. Check only single thing at one time, so you get full picture on what’s working and what’s not. And from there you can eventually isolate the problem and fix it.

It’ll always ask for the default configuration after hard reset. So, no, you won’t lose it forever.

That’s better, but you still need to have single wire to loop it around, which is not normally accessible. And at least in here the term ‘multimeter’ spesifically means one without a clamp, so you’d need to wire the multimeter in series with the load and that can be very dangerous if you don’t know what you’re doing.

Also, cheap ones often are not properly insulated nor rated for wall power (regardless of your voltage), so, again, if you don’t know what you are doing DO NOT measure current from a wall outlet with a multimeter.

or multimeter

Unless you really know what you’re doing DO NOT measure current on a wall outlet with a multimeter. Specially not with a cheap one. That can pretty easily break your hardware, burn your house down and kill you, not necessarily in that order.

“Enough battery life” is a bit wide requirement. What you’re running from that?

Most of the ‘big brands’ (eaton, apc…) work just fine with linux/open source, but specially low end consumer models even from big players might not and not all of them have any kind of port for data transfer at all.

Personally I’d say that if you’re looking for something smaller than 1000VA just get a brand new one. Bigger than those might be worth to buy used and just replace batteries, but that varies a lot. I’ve got few dirt cheap units around which apprently fried their charging circuit when the original battery died, so they’re e-waste now and on the other hand I have 1500VA cheap(ish) FSP which is running on 3rd or 4th set of batteries, so there’s not a definitive answer on what to get.

it might elude you how ridiculously fucking expensive stuff gets

I don’t know how that scales to worldwide CDN setup, but I work with a company who has presense on multiple countries and they drop casually 150-200k to few servers alone which provide services for at maximum for couple of thousand users. Networking, labor, power and things like that not included, just the hardware in a cardboard boxes.

Obviously there’s a ton of factors on this and I can’t elaborate our setup any further for obvious reasons, but just to give some scale how expensive things can get I can share my very real world experience. And that setup is pretty much the low end on the spectrum. 10k a month for a any as-a-service setup is almost a rounding error even with relatively low user count.

Serving anything to 100+ million people globally is a whole another beast. Wikipedia is a decent comparison, running a single website which looks like relatively simple to the end user (and oh boy it is not, but your Joe Average doesn’t know nor care about that) takes around 170 million dollars per year just on operating costs.

I don’t have any kind of opionion if that 30% is reasonable, but I do know that running that beast is not cheap and the money needs to come from somewhere. And as a customer, Steam just offers me what I want in a package which is the best one around, so they’ll keep getting my pennies until something better comes along. And I’m very aware that their service is not perfect and that I don’t really own anything on their platform, but for me in the current state in my life, they just provide the best bang for my buck and for the very limited time I have to spend with gaming.

Oh, I forgot to mention, I’m going to run the whole thing on a VPS, so I’ll have access to proper reverse dns and all, so that’s not an issue.

The couple of domains have been running on my company so the userbase is already there, but as I’m shutting down the business side they need to move to something else. And I don’t really care if users want to switch to something else, that’s not my problem.

All of those are still standing on Firefox’s shoulders and the actual rendering engine on the browser isn’t really trivial thing to build. Sure, they’re not going away, and likely Firefox will be around too for quite a while, but the world wide web as we currently know it is changing and Google and Microsoft are few of the bigger players pushing the change.

If you’re old enough you’ll remember the banners 'Best viewed with on ', and it’s not too far off from the future we’ll have if the big players get their wishes. Things like google suite, whatever meta is offering and pretty much “the internet” as your Joe Average understands it wants to implement technology where it’s not possible to block ads or modify the content you’re shown in any other way. It’s not too far off from your online banking and other very much real life affecting services start to have boundaries in place where they require certain level of ‘security’ from your browser and you can bet that things which allow content modifying things, like adblocker, doesn’t qualify for the new standards.

On many places it’s already illegal to modify or tamper DRM protected content in any ways (does anyone remember libdvdcss?) and the plan is to include similar (more or less) restrictions to the whole world wide web, which would say that we’ll have things like fediverse who allow browsers like firefox and ‘the rest’ like banking, flight/ticket/hotel/whatever booking sites, big news outlets and so on who only allow the ‘secure’ version of the browser. And that of course has very little to do with actual security, they just want control over your device and what content is fed to you, regardless if you like it or not.

I have no idea about cozy.io, but just to offer another option, I’ve been running Seafile for years and it’s pretty solid piece of hardware. And while it does have other stuff than just file storage/sharing, it’s mostly about just files and nothing else. Android client isn’t the best one around, but gets the job done (background tasks at least on mine tend to freeze now and then), on desktop it just works.