nixos/hosts/FredOS-Mediaserver.nix

{ config, pkgs, lib, ... }:
{
  config = lib.mkIf (config.networking.hostName == "FredOS-Mediaserver") {
    # Create symlink from home to storage
    systemd.tmpfiles.rules = [
      "L+ /home/fred/storage - - - - /mnt/storage"
    ];

    # Default max-jobs = 56 × cores = all let a from-source build thrash 31G
    # into a 10h livelock, taking routing down with it. Cap parallelism and
    # fence builds in a cgroup so the kernel OOM-kills a compiler, not the LAN.
    nix.settings.max-jobs = 8;
    nix.settings.cores = 8;
    nix.daemonCPUSchedPolicy = "batch";
    systemd.services.nix-daemon.serviceConfig = {
      MemoryHigh = "8G";
      MemoryMax = "10G";
      OOMPolicy = "continue"; # a killed build shouldn't stop the daemon
    };

    # System-wide backstop for whatever else runs away. oomd was running but
    # monitoring nothing; now it kills the worst cgroup once memory stalls
    # hit 80% for 30s (start of a thrash, not 10h in).
    systemd.oomd.enableRootSlice = true;
    systemd.oomd.enableSystemSlice = true;
    # Keep 1G truly free (default ~88M) so kernel networking and the watchdog
    # can still allocate while userspace is squeezed.
    boot.kernel.sysctl."vm.min_free_kbytes" = 1048576;

    # FredOS-Eee dials in and holds localhost:2222 -> its sshd open (see
    # hosts/FredOS-Eee.nix). Its key may open exactly that forward, nothing else.
    users.users.fred.openssh.authorizedKeys.keys = [
      ''restrict,port-forwarding,permitlisten="localhost:2222",command="${pkgs.coreutils}/bin/false" ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIDytimrM126GOrVlroh048Gt/w6Yv2x+Dp4rtUEueoYr eee-tunnel''
    ];
    # Reap dead sessions in ~3min so a dropped tunnel's port frees up for the
    # reconnect (default is TCP keepalive: 2h).
    services.openssh.settings.ClientAliveInterval = 60;
    services.openssh.settings.ClientAliveCountMax = 3;
    programs.ssh.extraConfig = ''
      Host eee
        HostName localhost
        Port 2222
        HostKeyAlias FredOS-Eee
        StrictHostKeyChecking accept-new
    '';

    # Basic system packages
    environment.systemPackages = with pkgs; [
      mergerfs
      wget
      util-linux
      # This host is the router (services/router.nix) but shipped without any
      # way to see a packet, so "is that nftables rule actually matching?" was
      # unanswerable. Counters tell you a rule fired; these tell you what it
      # fired on.
      tcpdump
      conntrack-tools
      javaPackages.compiler.temurin-bin.jre-25
      unzip
      screen
      # The Eee can't evaluate the flake in 1GB: build here, push over its
      # tunnel, apply on next boot (a live switch froze it out of memory).
      # The out-link keeps the closure from the weekly GC.
      (pkgs.writeShellScriptBin "eee-update" ''
        set -euo pipefail
        sys=$(nix build --refresh --print-out-paths --out-link ~/eee-system \
          git+https://forg.gregersen.it/rope/nixos#nixosConfigurations.FredOS-Eee.config.system.build.toplevel)
        nix copy --no-check-sigs --to ssh-ng://root@eee "$sys"
        ssh root@eee "nix-env -p /nix/var/nix/profiles/system --set $sys && $sys/bin/switch-to-configuration boot"
        echo "Eee updated; applies on its next reboot."
      '')
      yt-dlp
      # Audible library -> /mnt/storage/torrents/audiobooks for Jellyfin.
      # audible-cli pulls the .aaxc + voucher, AAXtoMP3 decrypts to m4b.
      audible-cli
      aaxtomp3
      ghostty.terminfo
      usbutils
      lm_sensors
      (pkgs.writeShellScriptBin "transcode-hevc" ''
        export PATH="${pkgs.jellyfin-ffmpeg}/bin:${pkgs.coreutils}/bin:${pkgs.findutils}/bin:${pkgs.gnugrep}/bin:${pkgs.gawk}/bin:${pkgs.bc}/bin:${pkgs.curl}/bin:$PATH"
        exec ${pkgs.bash}/bin/bash ${../scripts/transcode-hevc.sh} "$@"
      '')
      # Stats stream for the quickshell server monitor on the desktops:
      # `ssh mediaserver qs-stats` = top's 2s batch stream, interleaved with
      # one @STAT line per tick (hottest coretemp across both sockets, WAN
      # byte rates from eno1). Everything rides one SSH connection.
      (pkgs.writeShellScriptBin "qs-stats" ''
        # Single writer: the loop re-emits top's lines itself and injects a
        # @STAT line at each frame header. top writing the pipe directly in
        # parallel raced the injected lines (pipe writes aren't line-atomic)
        # and spliced @STAT mid-frame, where the client parser never saw it.
        C=${pkgs.coreutils}/bin
        export LC_ALL=C
        prev_rx=""
        ${pkgs.procps}/bin/top -b -d 2 -w 512 | while IFS= read -r line; do
          printf '%s\n' "$line"
          case $line in
            "top - "*)
              t=0
              for h in /sys/class/hwmon/*; do
                [ "$($C/cat "$h/name" 2>/dev/null)" = coretemp ] || continue
                for f in "$h"/temp*_input; do
                  v=$($C/cat "$f" 2>/dev/null || echo 0)
                  [ "$v" -gt "$t" ] && t=$v
                done
              done
              read -r rx tx < <(${pkgs.gawk}/bin/awk '$1 == "eno1:" {print $2, $10}' /proc/net/dev)
              if [ -n "$prev_rx" ]; then
                printf '@STAT temp=%s rxbps=%s txbps=%s\n' "$((t / 1000))" "$(( (rx - prev_rx) / 2 ))" "$(( (tx - prev_tx) / 2 ))"
              fi
              prev_rx=$rx
              prev_tx=$tx
              ;;
          esac
        done
      '')
      # Instant-answer backend for the quickshell launcher on the desktops:
      # `printf '%s' "question" | ssh mediaserver qs-ask`.
      #
      # Runs through Codex CLI using fred's ChatGPT subscription. Sign in on
      # the server as fred with `codex login` once; no API key is needed.
      # The launcher shows Wikipedia first and replaces it when this answers.
      #
      # Reads the question from stdin — passing it as an ssh argv element would
      # send it through the remote shell for a second round of word splitting.
      #
      # Luna at low reasoning effort is enough for a short fact. The CLI runs
      # read-only, without web search, shell access, user config, exec rules,
      # or session persistence.
      #
      # STDOUT is the launcher's channel: the answer, or nothing at all if the
      # call fails or the model doesn't know — silence means "keep Wikipedia".
      # Reasons go to STDERR and the journal (`journalctl -t qs-ask`).
      (pkgs.writeShellScriptBin "qs-ask" ''
        # Every line carries this script's pid, because the launcher fires a
        # fresh call per debounce and concurrent runs otherwise interleave with
        # no way to pair a question with its outcome. The pid goes in the
        # message text rather than through logger's --id: --id sets SYSLOG_PID,
        # which journalctl does not display — it shows the pid of the logger
        # process itself, a different number on every line.
        log() {
          printf 'qs-ask: %s\n' "$*" >&2
          ${pkgs.util-linux}/bin/logger -t qs-ask -- "[$$] $*"
        }

        q=$(${pkgs.coreutils}/bin/cat)
        [ -z "$q" ] && { log "no question on stdin"; exit 0; }
        log "ask: $q"

        # Keep repository instructions and project files out of this invocation.
        work="$HOME/.cache/qs-ask"
        ${pkgs.coreutils}/bin/mkdir -p "$work" || { log "cannot create $work"; exit 0; }
        cd "$work" || { log "cannot enter $work"; exit 0; }

        ans=$(printf 'Answer the question in one or two short sentences, under 240 characters. Lead with the specific fact asked for, including units. The query is often bare keywords rather than a question ("shadow hunter knife classic wow"); treat that as a request for the key facts about that thing. No preamble, caveats, markdown, or follow-up offers. Do not use tools. Reply with exactly UNKNOWN only when you genuinely do not know, or the answer depends on live data you do not have.\n\nQuestion: %s\n' "$q" | \
          ${pkgs.coreutils}/bin/timeout 30s ${pkgs.codex}/bin/codex exec \
          --model gpt-5.6-luna \
          --config model_reasoning_effort=low \
          --config web_search=disabled \
          --disable shell_tool \
          --sandbox read-only \
          --skip-git-repo-check \
          --ephemeral \
          --ignore-user-config \
          --ignore-rules \
          - \
          2>/dev/null) \
          || { log "codex exited non-zero (check 'codex login status' as fred)"; exit 0; }

        ans=$(printf '%s' "$ans" | ${pkgs.coreutils}/bin/tr -d '\r' | ${pkgs.gnused}/bin/sed -e 's/^[[:space:]]*//' -e 's/[[:space:]]*$//')

        case "$ans" in
          "")        log "empty answer"; exit 0 ;;
          UNKNOWN*)  log "model replied UNKNOWN"; exit 0 ;;
        esac
        log "ok: $ans"
        printf '%s' "$ans"
      '')
    ];

    # Basic networking
    networking.useDHCP = lib.mkForce false;

    # Allow fred to act as a remote Nix builder (trusted users can import
    # unsigned store paths sent by the build client).
    nix.settings.trusted-users = [ "root" "fred" ];

    # services/deployment.nix owns the 05:15 update and conditional reboot,
    # using the same health-checked transaction as manual and router UI applies.

    # WAN exposure is controlled by nftables in services/router.nix +
    # ports.toml (networking.firewall is disabled on this host).
    services.openssh = {
      enable = true;
      settings = {
        PermitRootLogin = "no";
        PasswordAuthentication = false;
      };
    };
  };
}