Frontend Services


I know, I’m stating the obvious in writing “frontend service allow you to do things.” If you are here, likely you know that too. So here’s a list of services I’ve employed to allow me to do a variety of, yep, things. I will add that most are setup as Docker containers across a few devices.


Audiobookshelf - Self-hosted audiobook and podcast server

Notes:

  • STATUS: Running and actively use it across mobile and desktop environments
  • LIKES: Supports both audiobooks & podcasts, auto downloads new podcasts & keeps track of play position across devices. Hosting the content frees up device space on the episodes I want to keep.



Calibre-Web - Self-hosted ebook server

Notes:

  • STATUS: Not a daily driver, but handy to have available.
  • LIKES: Any book in my digital library is only seconds away.
  • DISLIKES: Feature set is a bit stale, project largely in maintenance mode



Home Assistant - Home automation

See dedicated section on this topic.

Notes:

  • STATUS: It’s essential in my household at this point
  • LIKES: Open source and extremely customizable.
  • DISLIKES: None really, but there is a steep learning curve and it’s not for everybody.



Grafana - Dashboards & analytics

Notes:

  • STATUS: Running but not used frequently, report queries need maintenance
  • LIKES: Nice looking data visualizations.



Immich - Photo management and sharing server

Notes:

  • STATUS: Also not a daily driver but a very capable tool to manage a photo library spanning multiple years across multiple family devices.
  • LIKES: AI facial recognition, file geo-tag and mapping, album and video support
  • DISLIKES: Finer grain config settings would be helpful as AI processing can bring a low powered hosting device to a crawl.



Kanboard - Kanban style project management



Nextcloud - Dropbox like functionality, collaboration, & more

  • Current use: sync Obsidian files, ebook settings, file & photo sharing
    • Not actively using. Found other specialized services for my use cases + tired of too many maintenance headaches with each release.
  • Main: https://nextcloud.com/
  • Repository: https://github.com/nextcloud/server
  • Issue Resolution:
    • Steps to resolve DB connection issues on new setups (as of 10/01/19)
      1. Let the initial sign-in/setup process create the nextcloud user - don’t attempt to manually create this in the DB beforehand
      2. Once data has been entered into sign-in/setup page, the page will attempt to reload …..it may time-out at this step
      3. If a time-out is encountered from step 2, go into postgres and manually grant db permissions on the “OC_” prefixed user that was created during step 1
      4. Reload the main nextcloud url…..this should now take you to the login page instead of the setup. (note: the browser cache may need to be cleared beforehand).
    • When Nextcloud is behind a reverse proxy (as of 06/01/20 tested with Traefik v2)
      • Will not be able to connect properly without doing the following:
        • Include: - TRUSTED_PROXIES=xxx.xxx.xxx.xxx entry in docker compose
        • Add 'overwriteprotocol' => 'https', to config.php file
          • config.php location: in Docker persistant data …/nextcloud/www/config/
      • Error Msg: Cannot write into “config” directory!
  • Other Potentially Useful Links:

Notes:

  • STATUS: Setup but not actively using
  • LIKES: Easy way to remotely share files
  • DISLIKES: Too much feature bloat and constant breaking changes - cost to benefit ratio became lopsided and stopped using … maybe better now??



Node-RED - Home automation

See dedicated section on this topic.

Notes:

  • STATUS: NR is an essential service as it is effectively the automation backend for my home assistant instance.
  • LIKES: Automations are quick and easy to construct once you are familiar with the visual paradigm.



PaperlessNGX - Searchable document management system

Notes:

  • GOAL: Enter search terms and return list of pdf documents that contain said terms.
  • STATUS: Seems like paperNGX can do this, not sure how well yet. Need more time to assess… a local LLM with RAG capabilities may be able to do it better?