[prometheus] update guide for some reasons

[noave]

• [ ] guide needs to be updated to follow contributing guideline:

Always use full paths in commands. Don't assume the home directory or the html folder.

• [ ] setup daemon section is a bit strange, the port is wrong in example and one should immediately use 0.0.0.0, use highlighting changes instead of extra example
• [ ] use startsecs in config
• [ ] add backup, update, troubleshooting, logs
• [ ] cleanup and remove the install folder
• [ ] add workaround for htaccess proxy to use httpd

[noave]

@temal- as the guide author maybe you are interested to work on this :)

[maltekrupa]

Sure, I'll have a look at it.

[maltekrupa]

Hey,

setup daemon section is a bit strange, the port is wrong in example

it really is strange. Nice catch. 👍

one should immediately use 0.0.0.0

Having it listen on localhost and using a publicly reachable grafana on the same uberspace account was the initial target (for me). Can you elaborate why you think this is the way to go?

Doing backups via snapshots requires one to enable the admin API endpoints which really shouldn't be available publicly. But maybe a different backup approach (stop service, copy data, start service) could allow it to be reachable publicly.

use startsecs in config

I'll try to figure out a decent number for this, but since this is related to the size of the database (bigger database => longer startup time) it might not be that helpful. Tbh, I haven't worked with supervisord since creating this guide; maybe you have some insight to share? A health check would be nice to have.

add workaround for htaccess proxy to use httpd add backup, update, troubleshooting, logs

Haven't had time to look at these yet.

Malte