Selfhosted

I call an api in technitium to register dns for all services when they are instantiated, and route everything through an nginx reverse proxy - Sonarr.internal.tld for example

I don’t use any kind of monitoring Or dashboards

Set of cron jobs that check services, then send a Matrix message if there's an issue.

For the cron jobs, I pipe stderr to another script that watches those and does the same.

If all fails, and internet is unavailable and the router crashes, a Pi will toggle a relay, cutting and resupplying power.

If Plex doesn’t add new shows/movies/music then I take a look at my services that should be adding stuff for Plex to serve up. That’s pretty much it these days. I had a few pinned tabs in my browser for some of them so can see if they aren’t working if I click on them to add/change something.

I was using homepage but it seems to cause docker to die a LOT on my server.

I mucked around with so many dashboards, homarr, homepage, dashy but settled on glance

Mainly because it's minimalist and mostly text based. Handles my RSS feeds and anything that I want render I can usually vibe using the custom API widget.

https://github.com/glanceapp/glance

i am un-admining. free-range artisanal services wherever i happen to drop them. hell i don't even know what's running and what's not until i try to access something.

i manage tech all day so my home tech is nothing but abject chaos and i'm ok with that. i have backups and i can go without if needed.

service still up = no problem
Can't access service = problem, better ssh in

Simple as

Uptime Kuma monitoring anything I care about and notifying me via Matrix, or notifying me via email if it's Matrix that's down.

If something goes down my kids will be a more immediate and annoying alerting tool than anything I’ve used professionally.

I have just reduced the number of services to the couple I actually use, which I mostly remember exist. I have my own domain, so each service is service.mydomain.tld

I'm not, really. I run docker-compose and it runs. That's it.

If I had time to make dashboards, I wouldn’t waste it making dashboards. Most of the stuff I have just works without a lot of attention, and that’s the way I like it.

I just wait for someone to scream if it breaks.

Users, monitoring your services for free since internet exists

I'll notice it's down when I try to access it and it doesn't work. If it's not down, there is nothing to manage 🙃

I have documentation if I need to see everything at a glance. I don't need a live-updating dashboard for that.

Unraid has a table of the docker containers.

I don't need metrics or stats. I wouldn't look at, or care about them anyway. Dashboards feel like tech enthusiast crap. Tech and resources for the sake of having tech. My services are to solve a problem, not look at metrics of.

I just simply dont monitor most things. I do have a few things such as low disk space and failed backups. They are just simple shell scripts that send me an ntfy message when there is a problem.

Does dockge count as a dashboard?

'Cause I use that to quickly check on what's running, what's stopped. Then I do most of my mainenance in a terminal, via SSH to the server.

I don't know how you guys function without some sort of visual. I will forget everything I'm running if it's not on a dashboard of some sort. That's not a maybe - it's guaranteed. Because it's happened before.

Kubernetes with

• helm
  • the Kubernetes version of compose files
• fluxcd
  • manages the helm releases
• renovate
  • scans my github kubernetes repo for dependencies and creates pull requests for updates

I tried portainer for a while, but it was almost useless to me, as I'd always end up in the command line anyway. So I dropped that and any other dashboard idea.

https://charity.wtf/2021/08/09/notes-on-the-perfidy-of-dashboards/

Graphs and stuff might be useful for doing capacity planning or observing some trends, but most likely you don't need either.

If you want to know when something is down (and you might not need to know), set up alerts. (And do it well, you should only receive "actionable" alerts. And after setting alerts, you should work on reducing how many actionable things you have to do.)

(I did set up Nagios to send graphs to Clickhouse, plotted by Grafana. But mostly because I wanted to learn a few things and... I was curious about network latencies and wanted to plan storage a bit long term. But I could live perfectly without those.)

I use portainer, not sure if that counts as a dashboard?

Arch packages. All services have systemd integration.

I monitor everything with xymon, I get emails when there's a problem. Works like a charm.

Bookmarks for linking to services. Grafana for graphs that I only look at if I am curious or looking into when a problem arises. I could use Uptime Kuma if I wanted a simpler solution or notifications.

I do have Dashboards in Grafana, but I only use them to look something up. I have Prometheus Alertmanager connected to a Matrix bot that sends me messages when something looks wrong.

https://github.com/nicolargo/glances

I have a dashboard as well (Homepage), but this is a nice look at system resource usage and what's running, at a glance.

Uptime-kuma emails me when services or critical LAN devices are unreachable for whatever reason.

dokploy.com

FreshRSS to keep track of as much as possible, along with Uptime Kuma and plain old bookmarks

docker stats

With that command, I get all the stats I need, no dashboard required.

I use my own dashboard as a links page, nagios to monitor all the running servers and service's. Nagios will post to pushover if there's an issue.

docker ps