Files
silo-server/internal/nodeconfig/watcher.go
T
a17529fd6f feat(config): live admin settings + truthful restart-required banner (#128)
* feat(nodeconfig): harden config watcher for integrated-mode use

- RequestReload(): non-blocking, coalescing reload nudge that runs on the
  poll goroutine, so concurrent requests can never swap a stale snapshot
  over a newer one (unlike ForceReload from request handlers)
- Skip OnChange callbacks when the reloaded config is deep-equal to the
  previous one, so the 60s poll doesn't fire rebuild/log callbacks on
  no-op reloads
- Add RedisURL to BootstrapOverrides; previously a reload clobbered an
  env-provided Redis URL in the live config
- Split reload into fetchSettings/applySettings and add unit tests

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(config): hot-reload config watcher in integrated mode

Start nodeconfig.Watcher in integrated/api mode (previously only proxy/
transcode worker modes hot-reloaded). Expose the live config to the API
and jellycompat routers via func-typed LiveConfig/OnConfigChange fields
with nil fallbacks to the startup snapshot, and wire the admin settings
update hook to RequestReload so same-process changes apply immediately
even without Redis.

No consumer reads the live config yet — conversions land separately.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(admin): truthful restart-required banner for settings saves

The settings UI showed 'restart required' after every save regardless of
the key. The backend now classifies each key via a central registry
(internal/config/restart_keys.go) and PUT /admin/settings/{key} reports
restart_required per key; useSettingsForm only raises the banner when a
saved key actually needs a restart (and keeps it raised until restart).

The registry is conservative: every currently startup-frozen key is
marked restart-required; subsequent hot-reload conversions shrink it.
Settings read live from the settings repo (branding, overlays, markers,
download.*, ...) default to no-restart. DownloadSettings/OverlaySettings
drop their hardcoded restartRequired={false} special-casing.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(logging): hot-reload server.log_level and server.log_quiet

Share one slog.LevelVar across the handler chain and make
logfilter.Handler's quiet-prefix list an atomic pointer shared with
WithAttrs/WithGroup clones (New previously returned the inner handler
unwrapped when the quiet list was empty, leaving nothing to update).
The integrated-mode config watcher now applies both settings live;
their keys leave the restart-required registry.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(auth): hot-reload access/refresh token expiries

JWTService stores expiries as atomics with a SetExpiries hook; all three
instances (main API, ABS compat, jellycompat) re-apply them on config
reload. Applies to newly issued tokens; outstanding tokens keep their
original expiry. The JWT secret stays fixed for the process lifetime.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(playback): read transcode config live at session start

The playback and stream handlers pull ffmpeg path / hwaccel / transcode
dir from the live config when starting a transcode or extracting
subtitles, instead of values frozen at router construction. Each session
snapshots the config once so its output dir and binary stay consistent.

Also fixes a real bug: playback.hw_device was parsed into the config but
never wired into the integrated-mode handler, so local transcodes always
ran with an empty HWDevice while transcode nodes honored it.

playback.transcode_dir leaves the restart-required registry (the handler
is its only consumer); ffmpeg_path/hw_accel stay restart-required until
scanner/chapterthumbs/audiobook consumers convert.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(jellycompat): read compat identity settings live per request

System/Auth handlers take a config provider instead of the startup
snapshot, so jellyfin_compat.public_url, .server_name, and
.emulated_server_version apply without restart. server_id stays
restart-required (generate-once, baked into the resource mapper), as do
the session-store TTLs.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(scanner,metadata,mdblist): hot-reload worker pools and API key

scanner.workers, matcher.workers/batch_size, metadata.cache_images, and
mdblist.api_key convert to atomic fields with setters wired to the
config watcher. Worker counts apply on the next scan/match cycle (the
loops read them per cycle); the MDBList key applies to the next request.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(ai): hot-reload AI connection, models, toggles, and quotas

The shared llm.Client holds its config behind an atomic pointer
(UpdateConfig; each request snapshots once), and the subtitle/metadata
AI services gain UpdateConfig plus setters on the translator (batching)
and Whisper transcriber (ffmpeg path, chunk seconds). The router derives
their configs from shared helpers used both at construction and in
OnConfigChange callbacks, re-evaluating the chat-only-gateway transcribe
guard on each reload and warning only when it newly fires.

Everything on the AI Services page now applies without restart except
ai.max_concurrent_jobs (fixed-capacity dispatch semaphore).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(playback): wire transcode_enabled; remove dead playback/scanner knobs

playback.transcode_enabled was parsed into the config but the resolver
always received a hardcoded true — the admin toggle did nothing. It now
reads the live config per playback start, so disabling transcodes
applies without restart.

Remove settings that were wired to nothing so 'save + restart' stops
pretending: playback.allow_hevc_encoding (resolver field never
assigned), playback.transcode_ahead_segments and
playback.segment_duration (parsed, never consumed — segment duration is
per-session from the client), scanner.file_removal_grace (DeleteMissing
is never called). UI fields removed and the config struct fields pruned
so they don't resurrect; YAML import still tolerates the legacy keys.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-06-10 19:25:07 -04:00

229 lines
6.8 KiB
Go

package nodeconfig
import (
"context"
"fmt"
"log/slog"
"reflect"
"sync"
"time"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/Silo-Server/silo-server/internal/cache"
"github.com/Silo-Server/silo-server/internal/config"
"github.com/Silo-Server/silo-server/internal/secret"
)
// BootstrapOverrides holds values from env/CLI that must survive config
// reloads. These are set once at startup and re-applied after every
// LoadFromDB call.
type BootstrapOverrides struct {
Listen string // from PORT env
Mode string // from MODE env
DatabaseURL string // from DATABASE_URL env
JFListen string // from JF_PORT env
RedisURL string // from REDIS_URL env
}
// Watcher watches for configuration changes in the database and
// automatically reloads the Config when changes are detected.
type Watcher struct {
mu sync.RWMutex
cfg *config.Config
pool *pgxpool.Pool
cipher *secret.Cipher
eventBus cache.EventBus
bootstrap BootstrapOverrides
onChange []func(old, updated *config.Config)
reloadCh chan struct{} // buffered(1), event bus writes here
}
// NewWatcher creates a new config watcher. Call Start to begin watching. The
// cipher decrypts sensitive server_settings values (read here via raw SQL)
// before they reach config.LoadFromDB, so a hot reload never feeds ciphertext
// into the live config (which would, e.g., break JWT validation).
func NewWatcher(pool *pgxpool.Pool, cipher *secret.Cipher, eventBus cache.EventBus, bootstrap BootstrapOverrides) *Watcher {
return &Watcher{
pool: pool,
cipher: cipher,
eventBus: eventBus,
bootstrap: bootstrap,
reloadCh: make(chan struct{}, 1),
}
}
// Config returns the current config. Safe for concurrent use.
// Returns nil if Start has not been called.
func (w *Watcher) Config() *config.Config {
w.mu.RLock()
defer w.mu.RUnlock()
return w.cfg
}
// OnChange registers a callback invoked after a config swap whose new value
// differs from the old one. The callback receives the old and new config.
// Safe to call before or after Start; callbacks registered after Start only
// see reloads that happen after registration.
func (w *Watcher) OnChange(fn func(old, updated *config.Config)) {
w.mu.Lock()
defer w.mu.Unlock()
w.onChange = append(w.onChange, fn)
}
// Start performs the initial config load from the database, subscribes to
// EventSettingsChanged on the admin channel, and starts the background
// poll goroutine. Returns an error if the initial load fails.
func (w *Watcher) Start(ctx context.Context) error {
if err := w.reload(ctx); err != nil {
return fmt.Errorf("initial config load: %w", err)
}
// Subscribe to settings change events for immediate reload.
if err := w.eventBus.Subscribe(ctx, cache.ChannelAdmin, func(event cache.Event) {
if event.Type == cache.EventSettingsChanged {
select {
case w.reloadCh <- struct{}{}:
default:
// Already pending — coalesce.
}
}
}); err != nil {
slog.Warn("config watcher: subscribe to admin channel failed, using poll-only mode", "error", err)
}
go w.poll(ctx)
return nil
}
// ForceReload triggers an immediate config reload from the database.
func (w *Watcher) ForceReload(ctx context.Context) error {
return w.reload(ctx)
}
// RequestReload asks the poll goroutine to reload soon. Non-blocking and
// coalescing — safe to call from request handlers. Unlike ForceReload, the
// reload runs on the poll goroutine, so concurrent requests can never swap a
// stale snapshot over a newer one.
func (w *Watcher) RequestReload() {
select {
case w.reloadCh <- struct{}{}:
default:
// Already pending — coalesce.
}
}
// SetConfigForTest sets the config directly without loading from DB.
// This is intended for use in tests only.
func (w *Watcher) SetConfigForTest(cfg *config.Config) {
w.mu.Lock()
defer w.mu.Unlock()
w.cfg = cfg
}
// reload fetches all settings from the database, builds a new Config,
// applies bootstrap overrides, and atomically swaps the config pointer.
func (w *Watcher) reload(ctx context.Context) error {
m, err := w.fetchSettings(ctx)
if err != nil {
return err
}
return w.applySettings(m)
}
// fetchSettings reads all server_settings rows and decrypts sensitive values.
func (w *Watcher) fetchSettings(ctx context.Context) (map[string]string, error) {
rows, err := w.pool.Query(ctx, "SELECT key, value FROM server_settings")
if err != nil {
return nil, fmt.Errorf("query server_settings: %w", err)
}
defer rows.Close()
m := make(map[string]string)
for rows.Next() {
var k, v string
if err := rows.Scan(&k, &v); err != nil {
return nil, fmt.Errorf("scan server_settings row: %w", err)
}
// Decrypt sensitive keys (read-path contract: legacy plaintext passes
// through, enc:v1: values decrypt, corrupt ciphertext errors) so
// LoadFromDB always sees plaintext.
decrypted, derr := w.cipher.DecryptIfEncrypted(v, secret.SettingsAAD(k))
if derr != nil {
return nil, fmt.Errorf("decrypt server_settings %q: %w", k, derr)
}
m[k] = decrypted
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("iterate server_settings: %w", err)
}
return m, nil
}
// applySettings builds a Config from a plaintext settings map, re-applies
// bootstrap overrides, swaps the config pointer, and notifies OnChange
// callbacks when the config actually changed.
func (w *Watcher) applySettings(m map[string]string) error {
newCfg, err := config.LoadFromDB(m)
if err != nil {
return fmt.Errorf("parse config: %w", err)
}
// Re-apply bootstrap overrides — these are immutable for the process lifetime.
if w.bootstrap.Listen != "" {
newCfg.Server.Listen = w.bootstrap.Listen
}
if w.bootstrap.Mode != "" {
newCfg.Server.Mode = w.bootstrap.Mode
}
if w.bootstrap.DatabaseURL != "" {
newCfg.Database.URL = w.bootstrap.DatabaseURL
}
if w.bootstrap.JFListen != "" {
newCfg.JellyfinCompat.Listen = w.bootstrap.JFListen
}
if w.bootstrap.RedisURL != "" {
newCfg.Redis.URL = w.bootstrap.RedisURL
}
w.mu.Lock()
old := w.cfg
w.cfg = newCfg
callbacks := make([]func(old, updated *config.Config), len(w.onChange))
copy(callbacks, w.onChange)
w.mu.Unlock()
// The poll path reloads every 60s regardless of whether anything changed;
// don't fire callbacks (which may rebuild clients or log) on no-op swaps.
if old != nil && reflect.DeepEqual(*old, *newCfg) {
return nil
}
for _, fn := range callbacks {
fn(old, newCfg)
}
return nil
}
// poll runs the background loop that reloads config on timer or event.
func (w *Watcher) poll(ctx context.Context) {
ticker := time.NewTicker(60 * time.Second)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
if err := w.reload(ctx); err != nil {
slog.Warn("config poll reload failed", "error", err)
}
case <-w.reloadCh:
if err := w.reload(ctx); err != nil {
slog.Warn("config event reload failed", "error", err)
}
}
}
}