Music & Minecraft Part 4: The Sniper Daemon, Streams Watchdog, and Self-Healing Telemetry

-

Music & Minecraft Part 4: The Sniper Daemon, Streams Watchdog, and Self-Healing Telemetry

To move beyond my old PowerShell scripts and solve the nightmare of transient stream death specifically the endless rotation of expired authentication tokens, dropped sockets, and dead YouTube live containers I designed a fully autonomous, data-driven Python fleet running natively alongside mpd on the Gigabyte Brix.

As a DBA, I built the system around three core architectural principles: separation of concerns, state persistence, and aggressive self-healing telemetry.

1. Decoupled Data Architecture (Master vs. Slave)

The system strictly separates content definition from stream execution:

  • Master Files (Source of Truth): High-level, human-readable configuration manifests containing station lineups, metadata, and routing directives. These files remain completely isolated from the active player.
  • Slave Playlists (Execution Output): Compiled .m3u playlists generated automatically by the Python fleet and consumed directly by mpd. If an output playlist gets corrupted or stale, it is simply re-compiled from the Master without losing configuration history.

2. Three-Tier Operational Strategy

To keep CPU and network overhead minimal while avoiding aggressive API rate-limiting or unnecessary invocations, the fleet divides streams into three operational bands:

  • Completely Stable Tier: Standard Icecast and Shoutcast internet radio streams (BBC, LBC, global stations) that run continuously without active polling.
  • Unstable / Dynamic Tier: Streams powered by yt-dlp (such as YouTube feeds and BBC Sounds) where underlying containers, manifests, or authentication tokens require dynamic resolution, as well as RSS feeds that remain stable overall while seamlessly picking up newly added episodes.
  • The High-Churn "RAPID" Tier (#BRIX-RAPID): Volatile, highly dynamic feeds (such as live Twitch broadcasts and rolling YouTube streams) that frequently drop offline or change stream URIs. This tier is the dedicated hunting ground for the Sniper and Watchdog daemons, actively tracking real-time online/offline states.

3. The Python Fleet & Persistence Engine

Four specialized Python scripts and background daemons form the active telemetry and compilation suite:

The Master Generator (refresh_streams.py)

Compiles Master definitions into active Slave .m3u playlists using resolvers like yt-dlp.

  • Disk-Read Persistence Patch (#phash): Before writing new files, the generator inspects existing Slave playlists on disk using unique stream hashes (#phash). This ensures that routine playlist regenerations never overwrite active state titles, online uptimes, or dropout timestamps.

The Streams Watchdog (Healing Volatile YouTube Channels)

One of the trickiest parts of running a terminal audio setup is dealing with 24/7 rolling YouTube streams—like continuous live news feeds (e.g., Sky News) or 24/7 synthwave and lo-fi music channels hosted on YouTube. While they claim to be "24/7," YouTube's underlying streaming tokens, container endpoints, and manifest URLs constantly shift behind the scenes, causing standard static links to break after a few hours.

The Streams Watchdog actively monitors these volatile URLs. Instead of letting them die silently into static or throw a fatal player error, the watchdog continuously re-resolves the live stream endpoints in the background, updating the active stream pointers on the fly so these pseudo-permanent YouTube channels actually behave like reliable, un-interruptible broadcast feeds.

The Sniper Daemon (stream_sniper.py)

Operates as a fast, lightweight background sentinel, continuously sweeping the #BRIX-RAPID frequency band to audit real-time carrier status (is_live).

  • Kill & Re-acquire: If a stream drops off the airwaves, the sniper instantly snips the dead carrier from the active list, routes playback to a local fallback audio track, and holds the position. The exact millisecond the carrier pulses live again, the sniper re-acquires the feed and patches it back into the live stream seamlessly.
  • Precision Temporal State Tracking:
    • 🟢 ONLINE State: Maintains persistent, active uptime tracking while streams are healthy.
    • 🔴 OFFLINE - HOLDING RX State: Stamps exact ISO 8601 dropout timestamps (e.g., // 🔴 OFFLINE - HOLDING RX (2026-10-21T17:48:00Z)) when a stream dies, freezing the timestamp on disk so dropout logs survive routine system refreshes.
    • 🟡 CARRIER - ACQUIRING SB State: Tags newly discovered or unverified streams until the sniper locks onto their true carrier frequency.

The MPD Watchdog & Audio Guardrail

A dedicated companion daemon running alongside mpd. If an underlying stream thread hangs, a network socket drops, or audio processing stutters, the watchdog automatically resets the player daemon state and resumes playback without requiring human intervention.

The Ultimate Home-Lab Audio Engine

What started as a £60 eBay Gigabyte Brix bought for a fun weekend project has evolved into a headless, enterprise-grade, self-healing broadcast matrix.

Zero app bloat. Zero smart-speaker frustration. Just pure, resilient terminal audio backed by custom Python telemetry, automated YouTube URL healing, and the finest sci-fi atmosphere in the galaxy.

Comments

Post a Comment

Popular posts from this blog

Learning How to Deploy a AlmaLinux VM With Packer (Part 1)

-
Learning How to Deploy a AlmaLinux VM With Packer I looked into the process for using Packer, and while I understand its purpose, I wanted to gain a clearer insight into how to effectively implement it for my needs. Packer is a tool that automates the creation of machine images across various platforms. It allows you to define a configuration file that specifies how to build your images, ensuring consistency and repeatability in deployments. However, as I explored several templates for cloud-based systems, I found myself somewhat perplexed by the complexity of it all. I learn best by doing, but I didn’t want to dive into writing a template from scratch. So, I asked myself, what would Captain Picard do to get this thing up and running? He’d ask the ship’s computer. And in this case, I believed he might also consult Commander Data for a bit of extra insight. Consequently, I sought help from our own artificial intelligences to streamline the process, drawing on the information I had gat...
Read more

Installing and Using Git on Windows

-
Image
Installing Git on Windows: My Usual Method Here’s a straightforward guide to installing Git on Windows: Download the Latest Version Go to Git’s official download page and download the latest version for Windows. Run the Installer as Administrator Right-click the downloaded installer and select “Run as administrator.” Follow the Installation Wizard Information Screen: Click “Next.” Select Destination Folder: Click “Next” (accept the default option). Select Components: add "Add a Git Bash to windows Terminal" then click “Next.” Select Start Menu Folder: Click “Next” (accept the default option). Choosing the Default Editor: Select Notepad++ or Notepad, then click “Next.” Adjusting the Name of the Initial Branch: Click “Next” (accept the default option). Adjusting Your PATH Environment: Click “Next” (accept the default option). Choosing SSH Executable: Click “Next” (accept the default option). Choosing HTTPS Transport Backend: Select “Use the native Windows Secu...
Read more

Setting Up a RADV and NAT64 Router with Packer and VirtualBox

-
Setting Up a RADV and NAT64 Router with Packer and VirtualBox In this post, I'll go through setting up the core functionalities to make the bare bones VM fully operational as an IPv6 router, including configuring RADV and NAT64. After that, I'll secure the VM, power it down, and run some final checks to ensure everything is working as expected. Once validated the build phase will be complete and will be setting up for the final step, packaging it for deployment. Prevent Premature OVF Export To ensure that the VM remains available for further customisation before being wrapped into an OVF, I adjusted the Packer HCL file. This change allows me to apply additional configurations directly within the guest environment. By keeping the VM in this state, I have more flexibility to refine the build process before creating the final image. Here's the adjustment I made to the HCL file: skip_export = true keep_registered = true The skip_export setting prevents ...
Read more