It took me too long to get everything working myself because people love to share shit exclusively in CLI format and look down at anyone who asks for YAML it seems, so I’m always glad to pass it on
(I can understand CLI, but the ADHD brain finds YAML much easier for documentation purposes and it surprises me how many people seem to disagree)
Ok, had my wife send me the file from my network
networks:
main-network:
name: ${COMPOSE_PROJECT_NAME}
attachable: true
ipam:
driver: default
config:
- subnet: configure
ip_range: this
gateway: yoself
services:
# Gluetun - <https://github.com/qdm12/gluetun>
gluetun:
image: qmcgaw/gluetun
container_name: gluetun
networks:
- main-network
cap_add:
- NET_ADMIN
environment:
- PUID=${PUID}
- PGID=${PGID}
- TZ=${TZ}
- VPN_SERVICE_PROVIDER=custom
- VPN_TYPE=wireguard
- VPN_PORT_FORWARDING=true
- VPN_PORT_FORWARDING_PROVIDER=protonvpn
- WIREGUARD_ADDRESSES=use your own
- WIREGUARD_ALLOWED_IPS=0.0.0.0/0
- WIREGUARD_PRIVATE_KEY=nope
- WIREGUARD_PUBLIC_KEY=69420
- WIREGUARD_DNS=
- VPN_ENDPOINT_PORT=
- VPN_ENDPOINT_IP=
volumes:
- ${DOAPPDAT}/gluetun:/gluetun
I left in the wireguard stuff without my details because for me Gluetun refused to work when setting the exact same info to wg0.conf, so I define it in my compose
Then, services that rely on gluetun go below and look like:
# qBittorrent - <https://hub.docker.com/r/linuxserver/qbittorrent>
qbittorrent:
container_name: qbittorrent
network_mode: container:gluetun
image: lscr.io/linuxserver/qbittorrent:latest
depends_on:
gluetun:
condition: service_healthy
restart: unless-stopped
Works perfectly when I run it through portainer
What works for me:
Networks first in docker-compose
Gluetun first in Services, uses the network I set for it and the stack
Everything else goes below it, relying on the gluetun CONTAINER (I plan to have another stack running gluetun for other reasons so having it check the service is a no go for me) to be running in a HEALTHY state
All are set to restart: unless-stopped except gluetun, which is never
The expected behaviour is that containers will always wait for gluetun to report that it’s healthy before trying again to restart. Should gluetun fail and crash for any reason it won’t reboot and potentially fuck itself up harder, and no services will be able to start because it’s not reporting healthy.
This works perfectly in portainer and should when running docker-compose up, but for me it took portainer to work. Saw someone somewhere mention it has some sort of priority handling override built into it that docker itself doesn’t, meaning it’s less likely to fuck that lind of thing up, but idk how true it is
I’ll see if I can remember to snag a couple snips of my YAML to make it more clear
My main 2 reasons for installing it both come from needing to restart services sometimes:
Portainer let me allow other people access to restarting specific containers that occasionally misbehave
Portainer lets me update and restart all of the containers running in my VPN stack without breaking. For some ungodly reason, even with dependency set and everything in docker-compose, a CLI reboot will basically always start a service or 2 before gluetun is actually advertising it’s in a healthy state and everything breaks. With portainer that doesn’t happen, with the exact same compose, and I don’t get why lol
Ok i get it, it’s best practice to do rushed releases without QA because users are the free testers.
You literally used the wrong version. As I stated: the app you’re talking about clearly states it does not have a stable release for the version of nextcloud you’re running.
They definitely had no way to know that their own app was incompatible
They knew, and told you, right on the app page
Idiot user who believed their newsletter "update now, hub 9 is the best thing ever
You said it, not me. I tried being nice but that really is what happened: you fell for what the marketing team wrote and skipped basic IT steps in doing so. Now, rather than just admit you made a mistake that a LOT of people have made (including me, I’m a fucking idiot too) you are whining and doing your best to me talk gymnastics this into you being a victim of something
How you managed to convince your IT department of anything with a knowledge that shallow and an attitude like that I’ll never know. Grow up.
then a bit of warning is suggested
Which was given by the app that gets broken by the update
Windows doesn’t tell you that upgrading to 11 will break x, y, and z that you have installed, you’re expected to go to the sites for those programs and check if they work. Same exact idea
The same company making both apps is never a guarantee that they’ll play nice day 1, for many reasons
I’ll repeat: learn from your mistake instead of blaming other people for your naivete. If an app is important and might break during an update of something: check the apps documentation to see if it supports said update
Literally just googled “nextcloud forms” and looked at their supported versions and whaddya know, it says right on that webpage that there’s no stable version for 30 yet, so safe bet would be that it wouldn’t properly work when upgrading:
There is a supported nightly build, though, so you could probably have tried that
It’s on you to look up what will break when you update, or to test and see what happens when you do. A major update page isn’t going to list all of the things that rely on it that break because that’s fucking unreasonable
I’m starting to see a pattern in those comments like “why did you wear a skirt that night? It looks like you asked for it…”
Cute victim mentality, but gross and insanely wrong comparison
Learn from your mistake and don’t update without testing next time, it’s 100% on whoever updates the production environment to make sure that shit isn’t broken for whatever reason before pushing it customer-side
It’s more like you bought a random white powder from your dealer without asking what it was and are now upset you almost died
Many of us also lived through the era where any 13 year old could steal Mommy’s credit card and rent a botnet for that ezpz
My MC server a decade ago was tiny and it still happened every few months when we banned some butthurt kid
My SYNOLOGY NAS that’s just shy of being used for too much stuff (96% CPU means I’ve got 4% left babyyyy) is, was, and will forever be: Senpai.
That way people (my wife and parents) notice it in the network list.
Mobile apps are significantly better than they were at the start of the reddit migration, to the point connect has replaced Boost for me even though boost finally caught up to the rest of the class
So better in general I guess
Been thinking about swapping away from Plex when I move to New hardware later this month
If I’m reading all this right sounds like Overseerr and Plex both can be replaced with this, as well as a unified app for the two?
If so I’m def getting this installed and testing it out this weekend, my goal for a while has been to get the rest of my family into my streaming ecosystem but having to use 2 whole apps (Plex and Overseerr on a website) is enough to scare them away despite using 6 apps currently