I use rustdesk for remote desktop. Screen sharing is usually on zoom as it’s what my workplace uses.

random credentials + password manager

Nginx handles more connections than Apache, given the same resources. HAProxy does not have web server functionality like the former two, so Nginx is the natural upgrade from Apache. Caddy is relatively new, I’m not sure how it compares other than being easier to set up.

I have a ~40TB HDD array and jellyfin is super fast

I hope it is. But OP has a single drive.

As someone who had jellyfin running on HDD and moved it to a solid state: it’s a night and day difference when loading things. The reduced storage capacity for the same cost is very much worth it IMO.

OP wants it to last, so I wouldn’t consider used hardware an option

that might work; I don’t know if you live in a remote area, but I’d also consider a coffee shop, library, university, or hotel lobby with wifi. You might be able to download it within an hour.

A 50GB download takes less than 12h on a 10Mbps internet. And I had a 10Mbps link 10 years ago in a third world country, so maybe check your options with your ISP. 50GB really should not be a problem nowadays.

I’d look into AV1 decoding benchmarks, regardless of NVIDIA vs AMD, as I’ve been using NVIDIA on Jellyfin for a while with no issues.

HEVC is not as relevant IMO, as it’s not available through browsers due to license restrictions (ffmpeg / mpv works fine), so I’d focus on AV1 capabilities, which is not available in many cards.

deleted by creator

deleted by creator

then you’d limit the existing network addresses using subnets, as suggested by another answer in that question

yes, a bind mount / bind volume is when a volume is explicitly mapped to a location in your local storage rather than managed by docker and likely owned by root.

https://docs.docker.com/compose/compose-file/compose-file-v3/#volumes

Bind mounts. Always use bind mounts for data you care about, otherwise the “managed by docker” volumes are fated to be forgotten.

It won’t be your file structure as the file tree is managed by paperless, but at least using bind mounts you can easily navigate files and back them up independently or docker and paperless.

you don’t even need to think about letsencrypt

Do you know if it’s just as friction-less to have a self signed cert up with Caddy for internal use? I was using Nginx PM recently and had the need to serve https but I can’t use letsencrypt because it’s not public-facing. Nginx PM only has letsencrypt as an option.

I wish there was a checkbox that just deployed a self-signed cert without bothering with the details (it’s 2024 ffs, HTTPS should be 1 click away, whether that’s self-signed or not).

huh? What happened? Who’s shitting on ARM?

At ~5GB per HOUR? I don’t think so

Bandwidth (disk and network) is just one metric. Could it be an increase in number of IOPS due to syncing several small files?

Nothing, go ahead.

if these are the only services one is self hosting, I can see that.

But I have around a dozen stacks atm and I never came across a situation that I wanted to trigger an *arr stack restart with Jellyfin’s. They’re pretty much unrelated and independent services from an operational view.