ShellMonkey

If I recall right off hand it’s 92 cores, 64+8+8+12 between them. Been looking to merge a couple of the smaller ones into the others but it’s not a small task.

Way too damn much, around 700w between 4 rack servers and associated gear

Seems a pretty solid use of space, if you’re gonna have all this paved space for cars anybow why not keep them cool and generate power at the same time?

Mostly Zabbix and my own email server. Way I figure if there’s something broken enough to take out the email internally I’m probably already dealing with a broader issue.

I use a different ‘type 1’ hypervisor, create a small handful of VMs and then have a mess of containers inside containers in a couple of those. There’s a distinct VM for email since that has so many special requirements that it was just easier, one lxc/incus container host for admin tools, one for user apps, one for piefed because all these fediverse apps are so busy with the constant exchanges it kind of needs its own host to not affect others…

Why is more a case of ease of administration frankly. Pretty much everything gets auto patched and backed up and it’s easy to put up a new service or restore things if I hose something up.

The structure though has been a years long evolution that started out as ‘I just want a NAS to share files’ which then needed a frontend to make it more user friendly, and another app to play videos, and why not organize different file types rather than just dump em in a folder…

For context, that thing is about 3 stories tall and the helmet is a big screen that you can record yourself on a nearby camera to put yourself on.

Borg is its own thing, just handy to have it built in and preconfigured to work. If you run VMs on your working host instead of running NC directly on the machine you could just back those up for pretty much the same effect, though it wouldn’t usually be a realtime backup.

The next cloud data is fine being remote, when I say realtime I guess it could be better said as time sensitive. A database needs to update something and read it back in the right sequence for it to work properly. Just retrieving a file doesn’t have those issues.

If you happened to use it in a way like shared collaboration of a document with realtime updates it might be an issue, but I expect in that case the revisions would hit the DB first and then be put to the file as a secondary operation. Never tried that myself though.

Without getting into things like a SAN it’s generally not a good plan to put anything that needs realtime access remote to the device processing it. DBs are constantly making updates and introducing the odd bit of latency or a disconnect can make for lots of issues. I also have a worker/storage split setup and for a while tried just running the VMs as a whole using remote storage to host the vdisk, it wasn’t pretty even with a separate network dedicated to it.

For NC particularly, if you use their official AIO docker setup there is a built in borg backup you can store remotely and it makes restoration pretty painless.

I expect there’s some kind of TOS rule about scraping and recording of streams. Making it opt-out allows Twitch to sell those streams to AI companies similar to how Reddit did after going public.

What do they do when all the endpoints die for lack of parts and nobody can use their bots?

Recalbox or similar are certainly a thing, but that doesn’t really help with GOG games so much as roms. Depends on what the end goal is.