3 ms·

> That’s taking the structural discipline from the skyscraper and applying it to a space where I had total freedom. Yeah, nah. When I take my learnings home wi

> That’s taking the structural discipline from the skyscraper and applying it to a space where I had total freedom. Yeah, nah. When I take my learnings home with me, it fails every time. Usually, the scale of work necessary to maintain an enterprise-grade system rapidly outgrows the time I can reasonably allocate to it. In other cases, I lose interest because it's boring corporate crap. I don't known how all of you "homelab" people put up with it. I have enough Linux boxes at work that demand too much care and feeding. The author has a good point but it really isn't a two-way street. The hobby stuff can feed into your career, but letting it go the other way is usually either counterproductive, or bad for your mental health. Don't tinker in your shed because you think it'll advance your career. You'll be disappointed. Sorry for the spoiler. Tinker in your shed because it makes you happy, and brings joy and meaning to your life. You'll be more productive and, in my experience, you'll actually be more likely to learn something useful for work.

heyethan 7mo ago
[dead]
Normal_gaussian 7mo ago
I have a professional 'homelab' and a personal 'homelab'. You're 100% right, they can be a time sink. The important bit is to make sure the time is setup not 'maintenance' time. The trick is twofold: if it isn't 'declare and deploy' don't run it. If it isn't in your backup/restore pipeline don't run it. Pfsense and Home assistant are huge pains in the ass. Everything else is easy breezy. Proxmox/pbs/truenas/talos/linstor/DRBD are all amazing. I'm thinking about ditching pfsense for tailscale/cloudflare tunnels, but it's not worth the time atm. I don't have a viable alternative for HA.
- zrail 7mo ago
  Out of curiosity, what makes Home Assistant a pain?
  - ryukoposting 7mo ago
    I would guess I/O. Your normal containers need network access and that's it. HA, depending on your setup, might want Bluetooth, a USB zigbee dongle, z-wave, etc etc
    - Normal_gaussian 6mo ago
      No, the I/O passthroughs are fine. Proxmox and HA are fairly great at keeping them stable. I have passthrough for WiFi/BT/USB zigbee/USB thread Of course it pins you to a singular proxmox host, not benefitting from proxmox HA, but that's the way the cookie crumbles with h/w. pfsense and home-assistant both claim to be declarative configs, which is technically true. However the config files are not well or effectively documented, and where there is documentation it typically relates to the GUI which diverges significantly in arrangement. Their configs are declarative in that they declare the way their internal processes are configured, not in the way that they should interact and appear to other services (networking people will find that statement very confusing). Both are effectively "Operating Systems" within operating systems, starting/stopping/configuring/managing other programs, home assistant is doing this to the nth degree. When you start them it is very hard to determine when they have actually started - particularly the bits you care about. Getting errors and logs out of them is painful. Updating configs and restarting has multiple routes, the longest of which is very long. Both are reasonable ways to get to grips with the problem areas they solve for; they are not optimal however.
  - Normal_gaussian 6mo ago
    My other comment goes into more detail relating specifically to my earlier comment. But a niggle with Home Assistant in particular: It appears to be a system designed to be debugging adjacent, with every intent to frustrate debugging.
noirscape 7mo ago
The trick is to not overengineer your hobby if you're only doing it to prove a point. ie. Yes, you could run a full on corporate CA, issue SSL certificates for your domains, manually rig up wireguard and run your own internal corporate VPN... or you just accept that your grand total of 1 concurrent user on an intranet is probably just better served by setting up Tailscale and a wildcard LE certificate so that the browser shuts up. (Which is still not great, but the argument over HTTPS and intranets is not for right now.) Same with other deployment tools like Docker - yes, there's a ton of fancy ways to do persistent storage for serverless setups but get real: you're throwing the source folder in /opt/ and you have exactly one drive on that server. Save yourself the pain and just bind mount it to somewhere on your filesystem. Being able to back the folder up just with cp/rsync/rclone/scp is a lot easier than fiddling with docker's ambiguous mess of overlay2 subfolders. Every overengineered decision of today is tomorrow's "goddammit I need to ssh into the server again for an unexpected edgecase".
bob1029 7mo ago
I'm actually grateful (today) for the lightning strike that nuked my old pile of servers at home. It freed me from the whole thing in one step. I was completely disabused of any notion that I had control over anything at that point. You might think you are protected with UPSes and what not, but nothing will stop the electromagnetic effects if it hits within a few feet. Every piece of copper is going to get lit up. No solution is 100% guaranteed here, but EC2 and snapshots is a hell of a lot more likely to survive a single event like that.

heyethan 7mo ago

[dead]

Normal_gaussian 7mo ago

I have a professional 'homelab' and a personal 'homelab'. You're 100% right, they can be a time sink. The important bit is to make sure the time is setup not 'maintenance' time. The trick is twofold: if it isn't 'declare and deploy' don't run it. If it isn't in your backup/restore pipeline don't run it. Pfsense and Home assistant are huge pains in the ass. Everything else is easy breezy. Proxmox/pbs/truenas/talos/linstor/DRBD are all amazing. I'm thinking about ditching pfsense for tailscale/cloudflare tunnels, but it's not worth the time atm. I don't have a viable alternative for HA.

zrail 7mo ago
Out of curiosity, what makes Home Assistant a pain?
- ryukoposting 7mo ago
  I would guess I/O. Your normal containers need network access and that's it. HA, depending on your setup, might want Bluetooth, a USB zigbee dongle, z-wave, etc etc
  - Normal_gaussian 6mo ago
    No, the I/O passthroughs are fine. Proxmox and HA are fairly great at keeping them stable. I have passthrough for WiFi/BT/USB zigbee/USB thread Of course it pins you to a singular proxmox host, not benefitting from proxmox HA, but that's the way the cookie crumbles with h/w. pfsense and home-assistant both claim to be declarative configs, which is technically true. However the config files are not well or effectively documented, and where there is documentation it typically relates to the GUI which diverges significantly in arrangement. Their configs are declarative in that they declare the way their internal processes are configured, not in the way that they should interact and appear to other services (networking people will find that statement very confusing). Both are effectively "Operating Systems" within operating systems, starting/stopping/configuring/managing other programs, home assistant is doing this to the nth degree. When you start them it is very hard to determine when they have actually started - particularly the bits you care about. Getting errors and logs out of them is painful. Updating configs and restarting has multiple routes, the longest of which is very long. Both are reasonable ways to get to grips with the problem areas they solve for; they are not optimal however.
- Normal_gaussian 6mo ago
  My other comment goes into more detail relating specifically to my earlier comment. But a niggle with Home Assistant in particular: It appears to be a system designed to be debugging adjacent, with every intent to frustrate debugging.

noirscape 7mo ago

The trick is to not overengineer your hobby if you're only doing it to prove a point. ie. Yes, you could run a full on corporate CA, issue SSL certificates for your domains, manually rig up wireguard and run your own internal corporate VPN... or you just accept that your grand total of 1 concurrent user on an intranet is probably just better served by setting up Tailscale and a wildcard LE certificate so that the browser shuts up. (Which is still not great, but the argument over HTTPS and intranets is not for right now.) Same with other deployment tools like Docker - yes, there's a ton of fancy ways to do persistent storage for serverless setups but get real: you're throwing the source folder in /opt/ and you have exactly one drive on that server. Save yourself the pain and just bind mount it to somewhere on your filesystem. Being able to back the folder up just with cp/rsync/rclone/scp is a lot easier than fiddling with docker's ambiguous mess of overlay2 subfolders. Every overengineered decision of today is tomorrow's "goddammit I need to ssh into the server again for an unexpected edgecase".

bob1029 7mo ago

I'm actually grateful (today) for the lightning strike that nuked my old pile of servers at home. It freed me from the whole thing in one step. I was completely disabused of any notion that I had control over anything at that point. You might think you are protected with UPSes and what not, but nothing will stop the electromagnetic effects if it hits within a few feet. Every piece of copper is going to get lit up. No solution is 100% guaranteed here, but EC2 and snapshots is a hell of a lot more likely to survive a single event like that.