VMs can also be live migrated to another server in the cluster with no downtime and backups don’t need to take the VM down to do their thing. If in the future you want to move to physical hardware, you can use something like Clonezilla to back it up (not needed often, but still, something to consider).
Both have their places, but those factors are the main ones that come into play of when I want to use a VM or LXC.
Sorry, I didn’t mean to insinuate you were being insulting!
“Don’t feel crazy/bad/dumb, I’ve had the same thing happen to me!” is a pretty common phrasing in my region to show sympathy and understanding and I thought that’s what you had meant (and it sounds like for your area, ‘pregnant’ serves the same general purpose!).
It’s always crazy how that happens sometimes and after weeks of banging your head, everything just ‘clicks’ when you’re exposed to the information in the way that works best for you!
Dude, don’t feel pregnant.
Context clues, I assumed this autocorrect was some variation of crazy/bad/dumb? :-D
Weirdness: My Authentik instance had a PostgreSQL upgrade prerequisite in order to update it.
I’d followed instructions 3-4 times completely unsuccessfully and had to keep reverting to backup.
So, I gave up for a couple weeks and left it be in order to get over my frustration.
Yesterday, I followed the instructions again. As far as I can tell, I did nothing different than I’d tried previously and it worked first try and then I was also able to upgrade Authentik.
NOTE: The instructions aren’t exactly difficult! So, I don’t see how I’d have gotten it wrong!
To me, that’s the purpose of a “homelab” not the purpose of self hosting. There’s a lot of overlap, but they’re not quite the same. Homelab has a goal of learning, but just self hosting doesn’t need to.
Yeah, I bought the 3100 to support them and regretted that decision, unfortunately, when it came time to replace I was in a time crunch like you and wasn’t able to run my backups though a translation and it was taking way too long to do it manually so I had to just load pfSense and load the backup.
If I ever buy new hardware and the old isn’t dead though, I’m definitely going to try and make the shift away from it.
From their webpage … sounds pretty cool:
Sway is a tiling Wayland compositor and a drop-in replacement for the i3 window manager for X11. It works with your existing i3 configuration and supports most of i3’s features, plus a few extras.
Sway allows you to arrange your application windows logically, rather than spatially. Windows are arranged into a grid by default which maximizes the efficiency of your screen and can be quickly manipulated using only the keyboard.
I just installed it and it’s working pretty well.
OIDC/SSO was easy to configure and I was able to do so before even signing in. I was able to proxy it with NPM quite easily too without needing to do anything special.
The only real problem I’m seeing so far is that if you have OIDC set up, there aren’t prompts to actually use it in the Android app and Firefox extensions and it still prompts for username and password instead. I got around that by creating an API key instead, but you wouldn’t think that’d be necessary.
I even imported all my Firefox bookmarks just to see how it’d handle it and it’s struggling, haha, but I think that’s likely going to be the AI auto tagging and my poor little Ollama server that’s only got a 1060 rather than it being a Hoarder issue, but linking it to the existing Ollama server was also quite easy!
Thanks for the share OP, I’ve tried putzing with Wallabag (didn’t like that they didn’t have SSO) and Linkwarden (couldn’t get it to work with NGINX or NPM), so this was refreshing with how easy it was to get up and running!
ETA: My primary usecase for this is going to just be shoving things I want to remember to look at on it rather than sending myself links to things constantly.
Things I think could be improved, but am not (yet?) annoyed enough by to even open an issue:
You’re Welcome! An extra safety measure might be to do a clone on all your repos to ensure you’ve got a local copy of them all and absolute worst case you’ll have a couple of levels of backup plans, but up until pretty recently they were pretty much the same app just re-skinned, so, I think you’ll be fine.
`
systemctl stop gitea.service
cd /home/git/
wget https://codeberg.org/forgejo/forgejo/releases/download/v8.0.3/forgejo-8.0.3-linux-amd64
mv forgejo-* gitea
chmod +x gitea
systemctl start gitea.service
I did it soon after the “split up” though, but it was super easy since they were still basically the same applications.
Make backups, update the above to use your paths and the new download link you should be good to go. Mine is in a VM , so I was willing to just YOLO and give it a go since I could easily roll back.
sorry for the formatting. on my phone and did my best!
That sounds accurate. I have all my devices assigned a specific IP address, based on their MAC address, but that’s only per-interface. The other interfaces aren’t aware of my assignments for each other.
If I connect my phone to my LAN SSID, it’ll get its assigned IP, but if I connect it to the NOT [network of things, no internet access] SSID, it’ll get assigned a new address out of the DHCP pool because I haven’t assigned it an IP on that interface, until I assign it an IP. But, which VLAN it’s connected to will determine which IP its getting, and it still requires me to know the passwords for each SSID.
I believe where you’re getting confused is that a some businesses (or homelabs) might use a RADIUS server which will be more like this: ONE_SINGLE_SSID-Broadcast -> Device connects -> RADIUS Server detects account/certificate/MAC -> RADIUS Server assigns interface -> Device connects to VLAN the RADIUS server granted it access to
So, in that scenario, if the ONLY thing that’s being used to validate the access is the devices MAC address, just changing the MAC address will effectively grant a completely different level of access with nothing else changing. Most people in a homelab (and even plenty of larger businesses) aren’t running the infrastructure to do this though, they’re just effectively connecting a VLAN to a port and then that port can only be used to connect to that VLAN. They’re doing the same with the WiFi SSIDs where each SSID connects directly to the VLAN.
Usually though, for places that are implementing the RADIUS server, they’ll also install a certificate on their devices and the certificate needs to be in place in order to get certain access otherwise the RADIUS server will authorize less permissive access or just won’t allow access at all. Or, it’ll also need a user to log in to gain additional access.
For wired, the company may also implement port locking where the port will only allow a certain amount of MAC addresses to connect (presumably one unless there is also a VOICE VLAN with a phone being used, in which case it’d be two) where if you change your MAC address (or connect a different device), the port will lock and won’t power POE devices and won’t allow connectivity until an admin clears the lock. It’s possible that they may have multiple VLANs allowed on the port and client side you can change VLANs, but, this isn’t typically done on all ports, usually only on trusted ports or ports that need the multiple VLANs (my VM server for instance has access to a port that’ll allow multiple VLANs and I just enter the tag I need when I create the VM). This would be similar to your WiFi scenario, the port with the WAP connected to it will have access to multiple VLANs and then those WLANs just connect to the VLAN that they’re assigned to.
TL;DR - Typically one wireless SSID connects to one VLAN and if you want to jump to the other VLAN you’d need to connect to the other SSID, so you still have the individual passwords protecting you. On wired, typically VLANs are assigned per port and you can’t jump between then, but where they aren’t, it should be in a planned way and not just every port having access to every VLAN. Bad implementations exist though, so, anything is possible.