Files
silo-server/internal/playback/planstore/postgres_test.go
T
854d07cf8f feat(playback): add protocol v3 planning and recovery (#398)
* docs(playback): plan protocol v3 server implementation

* docs(playback): incorporate protocol v3 review

* feat(playback): implement protocol v3 server

* fix(playback): persist empty route diagnostics

* feat(playback): harden protocol v3 HDR routing

* feat(playback): complete protocol v3 client contract

* fix(playback): harden protocol v3 recovery

* fix(playback): restore dovi_rpu strip filter for DV remuxes

The v3 work renamed the Dolby Vision strip recipe to a dovi_split=mode=bl
bitstream filter that does not exist in stock FFmpeg or jellyfin-ffmpeg;
the probe failed closed on every deployment, disabling the new validated
DV7-to-HDR10 route and regressing the previously working dovi_rpu=strip=1
remux path from main. Restore dovi_rpu across the probe, remux and HLS
copy arguments, and the recipe-card constant.

Also from review: validate the remux DV mode for every profile (garbage
modes on non-P7 sources silently no-opped), reject preserve mode for P7
outright (a base-layer-only remux cannot preserve dual-layer DV), tag
dvhe sample entries only for the explicit v3 preserve recipe so legacy
web/jellycompat remuxes keep their pre-v3 hev1 labeling, and honor the
token-frozen DV mode in the proxy remux path instead of legacy-auto.

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

* fix(playback): correct v3 planner policy and contract validation

Review fixes to the v3 planner and wire contracts:

- Bar Profile 7 sources from the non-strip progressive remux route: a
  base-layer-only remux can never deliver native dual-layer DV, so the
  planner no longer emits plans claiming validated Dolby Vision while
  the executed remux drops the enhancement layer.
- Accept the device-quirks feature flag from either capability location,
  matching every other dual-location feature check.
- Treat legacy hdr_unknown rows as HDR10 for HDR10-capable clients with
  a degradation warning instead of leaving them unplayable under v3.
- Honor bandwidth_cap_kbps as a hard ceiling in every quality mode and
  wire the previously dead Metered signal into conservative auto rungs.
- Degrade to the validated source-quality route instead of a terminal
  when only an implicit quality reduction demanded an unsupported
  transcode; explicit user-selected rungs keep terminal behavior.
- Bound inner capability lists and strings; compare attempt keys exactly
  instead of case-folded; make ParseTrackIDV3 strict about canonical
  numerics; accept dvdsub/pgssub/dvbsub aliases and stop promising
  burn-in for unknown subtitle codecs; probe every h264 encoder rather
  than requiring libx264; normalize the file-level bitrate fallback.
- Evaluate subtitle renderability against the engine each candidate
  route executes on, not always media3_direct.
- Pin the with-quirks attempt-key preimage arity in the cross-language
  fixture so the Kotlin client stays in lockstep.

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

* fix(playback): harden v3 control-plane reliability

Review fixes to the v3 session, store, and handler layer:

- Bound concurrent replans with a slot semaphore: each replan pins a
  pooled connection for its advisory lock while issuing further store
  queries from the same pool, so an unbounded recovery storm could turn
  every connection into a lock holder and deadlock the server.
- Make CompleteReplan a real compare-and-swap (base-revision predicate,
  ErrReplanSupersededV3) and map BeginReplan insert races to a replay
  instead of a raw unique violation.
- Fingerprint start requests (request_digest column): an attempt ID
  reused with different input is now a 409-style conflict rather than a
  silent replay, and both replay paths check session liveness so dead
  sessions surface as retryable terminals.
- Pre-delete expired attempt rows on SaveAttempt so a retry during the
  cleanup window cannot wedge on an unreachable conflict.
- Align the in-memory store's semantics with Postgres and add DB-backed
  planstore tests (SILO_TEST_DATABASE_URL), including a regression test
  inserting every route-event name against the real CHECK constraint.
- Session manager: v3 route-set updates own RemuxDVMode outright so a
  replan onto an SDR source clears a stale strip mode; replacement
  reservations survive unrelated legacy stream updates; replacement
  admission excludes the replaced session explicitly instead of
  decrementing totals it may no longer be part of; the admission CAS
  loop is bounded and decider errors are logged.
- Map transient store failures to 500s instead of terminal 404/403s;
  authorize route events via identity-only projections after the rate
  limiter; keep sanitized diagnostics deterministic.
- Merge the server-computed durable plan key into replan exclusions so
  unreproducible client history cannot re-select the failed route.
- Remap tracks only when the effective edition changes (a same-file
  replan no longer switches audio to a lookalike track) and remap
  ID-only subtitle selections on edition fallback.
- Cache the v3/shadow feature flags for five seconds instead of one
  settings SELECT per playback request; stop remote transports
  best-effort when the start call times out; carry dvm/tid claims and
  the transport-scoped job identity through the legacy audio-change
  re-mint; index playback_route_events(received_at) for the retention
  delete; run store maintenance for DB-less deployments too.

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

* fix(transcode): reap idle node jobs and gate WebVTT conversion

- Add an idle reaper to the transcode node: a job untouched by manifest
  or segment requests for ten minutes is closed and unregistered. After
  a v3 replan retires a transport ID, a stale in-flight stream token
  could resurrect the old job via reconstruct and encode to end-of-file
  for nobody; jobs waiting on readiness count registration as access
  and are never reaped mid-wait, and reaping keeps the recipe so a
  still-valid token reconstructs on the next hit.
- Reject bitmap subtitle tracks (PGS) on the .vtt conversion path with
  415 before headers are written instead of spawning an ffmpeg command
  that always fails mid-response, and make the extract-format override
  fall back to source-driven mapping for bitmap codecs.
- Drain error bodies on non-202 node responses so the HTTP transport
  can reuse connections.
- Pin the transcode-dir cleanup separator-boundary semantics with a
  regression test (a session ID sharing another's prefix must not
  retain foreign directories).

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

* fix(playback): close v3 planner policy gaps from review

- Clamp the final transcode bitrate to bandwidth_cap_kbps: the ladder has
  no rung below 480p/1500kbps, so lower caps were silently exceeded even
  though the cap is documented as a hard delivery ceiling.
- Treat video-only media as audio-compatible instead of forcing an AAC
  conversion (or an audio_conversion_unsupported terminal) onto a file
  with no audio stream. Tracks whose codec failed to probe keep the gate.
- Only promise a bitmap subtitle sidecar for embedded PGS with an engine
  that renders embedded bitmap: external/downloaded bitmap and embedded
  DVD/DVB published artifact URLs that always failed at fetch. They now
  fall through to burn-in or its terminal.
- Accept client_video_transformations_v1 from either client_features or
  the nested context when validating client-executor transformations,
  matching the planner's dual-source reads.

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

* fix(playback): probe and execute DV remuxes with one ffmpeg binary

The v3 transformation registry probed the configured playback.ffmpeg_path
while progressive remux execution resolved the process-global discovery
path, so a deployment where only one binary carries dovi_rpu could plan a
server_dv7_to_hdr10 route and then fail it at stream time. Resolution now
goes through a shared ResolveFFmpegPath (configured path first, discovery
fallback — the same rule the transcode pipeline already used), the
dovi_rpu probe is cached per binary path, and the stream handler and proxy
worker pass their configured path into ServeRemuxWithDVMode.

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

* fix(playback): harden v3 replan identity and control-plane limits

- Seed failure-replan track selections from the durable current plan
  before overlaying the request: after an alternate-version fallback the
  normalized request still carries requested-edition track IDs, so a
  replan omitting unchanged tracks was rejected as a track/file mismatch.
- Remap ID-only audio selections across edition changes (parse the ID to
  an index like the subtitle remap already does) instead of leaving a
  stale file-bound ID to fail validation.
- Release the node planner reservation when a prepared remote transport
  rolls back after the node accepted the job; repeated failed starts
  could otherwise pin max-job/bandwidth budgets for the full reservation
  age.
- Size the replan semaphore below the PostgreSQL pool via a store
  capacity advisor: with max_connections at or below the fixed bound,
  advisory-lock holders could starve the inner store queries they need
  to finish.
- Contain shadow-planner panics with a recover boundary; it runs on a
  bare goroutine where an escaped panic kills the process for what is
  telemetry-only work. Document why the memory store's session lock is
  deliberately a no-op.

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

* fix(transcode): serialize node job teardown against reconstructs

- Look up and touch manifest/segment sessions in one critical section so
  the idle reaper cannot unregister a job between the lookup and its
  liveness refresh.
- Re-validate each reap candidate under the per-session lifecycle lock
  before closing it: Close removes the output directory, and without the
  lock it could race a token reconstruct and wipe the segments the fresh
  ffmpeg is writing.
- Take the lifecycle lock in handleStop so a stop racing a RequireReady
  start's readiness wait blocks until registration and tears the job
  down, instead of 404ing and orphaning the ffmpeg until the reaper.

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

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-14 11:51:27 -04:00

602 lines
23 KiB
Go

package planstore
import (
"bytes"
"context"
"encoding/json"
"errors"
"fmt"
"os"
"testing"
"time"
"github.com/google/uuid"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/Silo-Server/silo-server/internal/playback"
)
// planstoreFixture holds the minimal FK graph (user + media files) that
// playback_v3_attempts and playback_route_events rows require.
type planstoreFixture struct {
pool *pgxpool.Pool
userID int
mediaFileID int
altFileID int
}
// newPlanstoreFixture connects to SILO_TEST_DATABASE_URL (skipping when
// unset), verifies the v3 migrations are applied, and inserts the fixture
// rows every attempt/event insert depends on. Cleanup deletes everything the
// tests wrote so reruns against the same database stay green.
func newPlanstoreFixture(t *testing.T) *planstoreFixture {
t.Helper()
dsn := os.Getenv("SILO_TEST_DATABASE_URL")
if dsn == "" {
t.Skip("SILO_TEST_DATABASE_URL is not set")
}
ctx := context.Background()
pool, err := pgxpool.New(ctx, dsn)
if err != nil {
t.Fatalf("connect test database: %v", err)
}
t.Cleanup(pool.Close)
var tableName *string
if err := pool.QueryRow(ctx, `SELECT to_regclass('public.playback_v3_attempts')::text`).Scan(&tableName); err != nil {
t.Fatalf("check playback_v3_attempts table: %v", err)
}
if tableName == nil || *tableName == "" {
t.Skip("test database has not applied the playback protocol v3 migration")
}
var hasRevision bool
if err := pool.QueryRow(ctx, `
SELECT EXISTS (
SELECT 1 FROM information_schema.columns
WHERE table_name = 'playback_v3_attempts' AND column_name = 'current_replan_request_id'
)`).Scan(&hasRevision); err != nil {
t.Fatalf("check current_replan_request_id column: %v", err)
}
if !hasRevision {
t.Skip("test database has not applied the playback v3 attempt revision migration")
}
f := &planstoreFixture{pool: pool}
unique := fmt.Sprintf("planstore-test-%d", time.Now().UnixNano())
var folderID int
if err := pool.QueryRow(ctx, `
INSERT INTO media_folders (type, name) VALUES ('movies', $1) RETURNING id`, unique).Scan(&folderID); err != nil {
t.Fatalf("insert fixture media folder: %v", err)
}
if err := pool.QueryRow(ctx, `
INSERT INTO users (username) VALUES ($1) RETURNING id`, unique).Scan(&f.userID); err != nil {
t.Fatalf("insert fixture user: %v", err)
}
if err := pool.QueryRow(ctx, `
INSERT INTO media_files (media_folder_id, file_path) VALUES ($1, $2) RETURNING id`,
folderID, "/fixtures/"+unique+"/movie.mkv").Scan(&f.mediaFileID); err != nil {
t.Fatalf("insert fixture media file: %v", err)
}
if err := pool.QueryRow(ctx, `
INSERT INTO media_files (media_folder_id, file_path) VALUES ($1, $2) RETURNING id`,
folderID, "/fixtures/"+unique+"/movie-alt.mkv").Scan(&f.altFileID); err != nil {
t.Fatalf("insert alternate fixture media file: %v", err)
}
t.Cleanup(func() {
cleanupCtx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
defer cancel()
// Attempts cascade to replans; deleting the user cascades any attempt
// or route event a failed subtest left behind; the folder cascades the
// media files.
_, _ = pool.Exec(cleanupCtx, `DELETE FROM playback_route_events WHERE user_id = $1`, f.userID)
_, _ = pool.Exec(cleanupCtx, `DELETE FROM playback_v3_attempts WHERE user_id = $1`, f.userID)
_, _ = pool.Exec(cleanupCtx, `DELETE FROM users WHERE id = $1`, f.userID)
_, _ = pool.Exec(cleanupCtx, `DELETE FROM media_folders WHERE id = $1`, folderID)
})
return f
}
func (f *planstoreFixture) attemptRecord(sessionID, attemptID, digest string) playback.AttemptRecordV3 {
return playback.AttemptRecordV3{
PlaybackAttemptID: attemptID,
SessionID: sessionID,
UserID: f.userID,
ProfileID: "profile-1",
RequestedMediaFileID: f.mediaFileID,
EffectiveMediaFileID: f.mediaFileID,
CurrentPlanID: "plan-1",
CurrentReplanRequestID: "",
CurrentPlan: playback.PlanV3{
ProtocolVersion: 3,
PlanID: "plan-1",
SessionID: sessionID,
DecisionReason: "direct_play",
RequestedMediaFileID: f.mediaFileID,
EffectiveMediaFileID: f.mediaFileID,
},
NormalizedRequest: playback.StartRequestV3{
ProtocolVersion: 3,
FileID: f.mediaFileID,
ProfileID: "profile-1",
PlaybackAttemptID: attemptID,
QualityPreference: "auto",
},
RequestDigest: digest,
ExpiresAt: time.Now().Add(time.Hour).UTC().Truncate(time.Microsecond),
}
}
func (f *planstoreFixture) expireAttempt(t *testing.T, attemptID string) {
t.Helper()
tag, err := f.pool.Exec(context.Background(), `
UPDATE playback_v3_attempts SET expires_at = NOW() - INTERVAL '1 minute'
WHERE playback_attempt_id = $1`, attemptID)
if err != nil {
t.Fatalf("expire attempt %s: %v", attemptID, err)
}
if tag.RowsAffected() != 1 {
t.Fatalf("expire attempt %s: affected %d rows", attemptID, tag.RowsAffected())
}
}
// mustJSON canonicalizes a value through encoding/json so structs that
// round-trip via JSONB can be compared without tripping on nil-vs-empty
// map/slice differences.
func mustJSON(t *testing.T, v any) []byte {
t.Helper()
data, err := json.Marshal(v)
if err != nil {
t.Fatalf("marshal: %v", err)
}
return data
}
func TestPostgresPlanStore(t *testing.T) {
f := newPlanstoreFixture(t)
store := NewPostgres(f.pool)
ctx := context.Background()
// Regression test for the CHECK-constraint drift: every event name the
// code emits must be accepted by the playback_route_events CHECK in the
// real schema.
t.Run("RecordRouteEventAcceptsAllEventNames", func(t *testing.T) {
sessionID := uuid.NewString()
names := playback.RouteEventNamesV3()
if len(names) == 0 {
t.Fatal("RouteEventNamesV3 returned no events")
}
for _, name := range names {
err := store.RecordRouteEvent(ctx, playback.RouteEventRecordV3{
RouteEventV3: playback.RouteEventV3{
ProtocolVersion: 3,
PlaybackAttemptID: "att-events-" + sessionID,
SessionID: sessionID,
PlanID: "plan-1",
PlanAttemptID: "plan-attempt-1",
PlanAttemptKey: "plan-attempt-key-1",
Event: name,
FailureClassification: "decode_error",
FallbackReason: "test",
OutputRouteGeneration: 1,
Diagnostics: map[string]string{"source": "planstore-test"},
},
UserID: f.userID,
ProfileID: "profile-1",
ClientName: "planstore-test",
ClientVersion: "1.0",
ClientModel: "test-model",
})
if err != nil {
t.Errorf("RecordRouteEvent(%q) rejected by real schema: %v", name, err)
}
}
var count int
if err := f.pool.QueryRow(ctx, `
SELECT COUNT(*) FROM playback_route_events WHERE session_id = $1::uuid`, sessionID).Scan(&count); err != nil {
t.Fatalf("count route events: %v", err)
}
if count != len(names) {
t.Fatalf("persisted %d route events, want %d", count, len(names))
}
})
t.Run("SaveAttemptIdempotency", func(t *testing.T) {
sessionID := uuid.NewString()
attemptID := "att-save-" + sessionID
record := f.attemptRecord(sessionID, attemptID, "digest-a")
if err := store.SaveAttempt(ctx, record); err != nil {
t.Fatalf("fresh SaveAttempt: %v", err)
}
// Exact replay of the same attempt-ID and digest.
if err := store.SaveAttempt(ctx, record); !errors.Is(err, playback.ErrPlaybackAttemptExistsV3) {
t.Fatalf("same-digest replay: got %v, want ErrPlaybackAttemptExistsV3", err)
}
// Same attempt-ID reused with different input (digest) is an
// idempotency violation, not a replay.
conflicting := f.attemptRecord(uuid.NewString(), attemptID, "digest-b")
if err := store.SaveAttempt(ctx, conflicting); !errors.Is(err, playback.ErrIdempotencyKeyReusedV3) {
t.Fatalf("different-digest reuse: got %v, want ErrIdempotencyKeyReusedV3", err)
}
// Once the original row expires, the pre-delete path must clear it so
// the attempt-ID becomes reusable.
f.expireAttempt(t, attemptID)
if err := store.SaveAttempt(ctx, record); err != nil {
t.Fatalf("SaveAttempt after expiry should reclaim the attempt-ID: %v", err)
}
})
t.Run("GetAttemptRoundTrip", func(t *testing.T) {
sessionID := uuid.NewString()
attemptID := "att-get-" + sessionID
record := f.attemptRecord(sessionID, attemptID, "digest-get")
if err := store.SaveAttempt(ctx, record); err != nil {
t.Fatalf("SaveAttempt: %v", err)
}
for name, fetch := range map[string]func() (*playback.AttemptRecordV3, error){
"GetAttempt": func() (*playback.AttemptRecordV3, error) { return store.GetAttempt(ctx, sessionID) },
"GetAttemptByPlaybackAttemptID": func() (*playback.AttemptRecordV3, error) { return store.GetAttemptByPlaybackAttemptID(ctx, attemptID) },
} {
got, err := fetch()
if err != nil {
t.Fatalf("%s: %v", name, err)
}
if got.PlaybackAttemptID != attemptID || got.SessionID != sessionID {
t.Fatalf("%s identity mismatch: %+v", name, got)
}
if got.UserID != record.UserID || got.ProfileID != record.ProfileID {
t.Fatalf("%s ownership mismatch: %+v", name, got)
}
if got.RequestedMediaFileID != record.RequestedMediaFileID || got.EffectiveMediaFileID != record.EffectiveMediaFileID {
t.Fatalf("%s media file mismatch: %+v", name, got)
}
if got.CurrentPlanID != record.CurrentPlanID || got.CurrentReplanRequestID != record.CurrentReplanRequestID {
t.Fatalf("%s plan revision mismatch: %+v", name, got)
}
if got.RequestDigest != record.RequestDigest {
t.Fatalf("%s request_digest = %q, want %q", name, got.RequestDigest, record.RequestDigest)
}
if !bytes.Equal(mustJSON(t, got.CurrentPlan), mustJSON(t, record.CurrentPlan)) {
t.Fatalf("%s plan JSON did not round-trip:\n got %s\nwant %s", name, mustJSON(t, got.CurrentPlan), mustJSON(t, record.CurrentPlan))
}
if !bytes.Equal(mustJSON(t, got.NormalizedRequest), mustJSON(t, record.NormalizedRequest)) {
t.Fatalf("%s normalized request JSON did not round-trip", name)
}
if diff := got.ExpiresAt.Sub(record.ExpiresAt); diff < -time.Millisecond || diff > time.Millisecond {
t.Fatalf("%s expires_at drifted by %v", name, diff)
}
}
identity, err := store.GetAttemptIdentity(ctx, sessionID)
if err != nil {
t.Fatalf("GetAttemptIdentity: %v", err)
}
byAttempt, err := store.GetAttemptIdentityByPlaybackAttemptID(ctx, attemptID)
if err != nil {
t.Fatalf("GetAttemptIdentityByPlaybackAttemptID: %v", err)
}
for name, got := range map[string]*playback.AttemptIdentityV3{"bySession": identity, "byAttempt": byAttempt} {
if got.PlaybackAttemptID != attemptID || got.SessionID != sessionID ||
got.UserID != f.userID || got.ProfileID != "profile-1" {
t.Fatalf("%s identity ownership mismatch: %+v", name, got)
}
}
// Expired rows must be invisible to every read path.
f.expireAttempt(t, attemptID)
if _, err := store.GetAttempt(ctx, sessionID); !errors.Is(err, playback.ErrSessionNotFound) {
t.Fatalf("GetAttempt on expired row: got %v, want ErrSessionNotFound", err)
}
if _, err := store.GetAttemptByPlaybackAttemptID(ctx, attemptID); !errors.Is(err, playback.ErrSessionNotFound) {
t.Fatalf("GetAttemptByPlaybackAttemptID on expired row: got %v, want ErrSessionNotFound", err)
}
if _, err := store.GetAttemptIdentity(ctx, sessionID); !errors.Is(err, playback.ErrSessionNotFound) {
t.Fatalf("GetAttemptIdentity on expired row: got %v, want ErrSessionNotFound", err)
}
if _, err := store.GetAttemptIdentityByPlaybackAttemptID(ctx, attemptID); !errors.Is(err, playback.ErrSessionNotFound) {
t.Fatalf("GetAttemptIdentityByPlaybackAttemptID on expired row: got %v, want ErrSessionNotFound", err)
}
})
t.Run("BeginReplanLifecycle", func(t *testing.T) {
sessionID := uuid.NewString()
attemptID := "att-replan-" + sessionID
if err := store.SaveAttempt(ctx, f.attemptRecord(sessionID, attemptID, "digest-replan")); err != nil {
t.Fatalf("SaveAttempt: %v", err)
}
future := time.Now().Add(time.Minute)
// New replan request: caller owns the lease.
lease, err := store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-1", "", future)
if err != nil {
t.Fatalf("BeginReplan new: %v", err)
}
if lease.State != playback.ReplanLeaseOwnedV3 {
t.Fatalf("BeginReplan new state = %q, want owned", lease.State)
}
// Same request-ID with different input.
if _, err := store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-other", "", future); !errors.Is(err, playback.ErrIdempotencyKeyReusedV3) {
t.Fatalf("digest mismatch: got %v, want ErrIdempotencyKeyReusedV3", err)
}
// Active unexpired lease held by someone else.
lease, err = store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-1", "", future)
if err != nil {
t.Fatalf("BeginReplan in-flight: %v", err)
}
if lease.State != playback.ReplanLeaseInFlightV3 {
t.Fatalf("BeginReplan in-flight state = %q, want in_flight", lease.State)
}
// Completed replan replays the stored response.
completed := f.attemptRecord(sessionID, attemptID, "digest-replan")
completed.CurrentPlanID = "plan-2"
completed.CurrentReplanRequestID = "rq-1"
completed.CurrentPlan.PlanID = "plan-2"
response := json.RawMessage(`{"plan_id": "plan-2"}`)
if err := store.CompleteReplan(ctx, sessionID, "rq-1", "", response, completed); err != nil {
t.Fatalf("CompleteReplan: %v", err)
}
lease, err = store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-1", "", future)
if err != nil {
t.Fatalf("BeginReplan completed: %v", err)
}
if lease.State != playback.ReplanLeaseCompletedV3 {
t.Fatalf("BeginReplan completed state = %q, want completed", lease.State)
}
var storedResponse, wantResponse any
if err := json.Unmarshal(lease.Response, &storedResponse); err != nil {
t.Fatalf("unmarshal replayed response: %v", err)
}
if err := json.Unmarshal(response, &wantResponse); err != nil {
t.Fatalf("unmarshal expected response: %v", err)
}
if !bytes.Equal(mustJSON(t, storedResponse), mustJSON(t, wantResponse)) {
t.Fatalf("replayed response = %s, want %s", lease.Response, response)
}
// Expired lease whose base revision no longer matches the retry.
past := time.Now().Add(-time.Minute)
if _, err := store.BeginReplan(ctx, sessionID, "rq-stale", "rq-digest-stale", "base-x", past); err != nil {
t.Fatalf("BeginReplan seed stale lease: %v", err)
}
if _, err := store.BeginReplan(ctx, sessionID, "rq-stale", "rq-digest-stale", "base-y", future); !errors.Is(err, playback.ErrStaleReplanLeaseV3) {
t.Fatalf("expired lease with stale base: got %v, want ErrStaleReplanLeaseV3", err)
}
// Expired lease with a matching base is re-owned.
if _, err := store.BeginReplan(ctx, sessionID, "rq-retry", "rq-digest-retry", "rq-1", past); err != nil {
t.Fatalf("BeginReplan seed expired lease: %v", err)
}
lease, err = store.BeginReplan(ctx, sessionID, "rq-retry", "rq-digest-retry", "rq-1", future)
if err != nil {
t.Fatalf("BeginReplan re-own expired lease: %v", err)
}
if lease.State != playback.ReplanLeaseOwnedV3 {
t.Fatalf("re-owned lease state = %q, want owned", lease.State)
}
})
t.Run("CompleteReplan", func(t *testing.T) {
sessionID := uuid.NewString()
attemptID := "att-complete-" + sessionID
if err := store.SaveAttempt(ctx, f.attemptRecord(sessionID, attemptID, "digest-complete")); err != nil {
t.Fatalf("SaveAttempt: %v", err)
}
future := time.Now().Add(time.Minute)
if _, err := store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-1", "", future); err != nil {
t.Fatalf("BeginReplan: %v", err)
}
updated := f.attemptRecord(sessionID, attemptID, "digest-complete")
updated.EffectiveMediaFileID = f.altFileID
updated.CurrentPlanID = "plan-2"
updated.CurrentReplanRequestID = "rq-1"
updated.CurrentPlan.PlanID = "plan-2"
updated.CurrentPlan.EffectiveMediaFileID = f.altFileID
updated.CurrentPlan.DecisionReason = "transcode_fallback"
updated.ExpiresAt = time.Now().Add(2 * time.Hour).UTC().Truncate(time.Microsecond)
response := json.RawMessage(`{"plan_id": "plan-2", "status": "replanned"}`)
if err := store.CompleteReplan(ctx, sessionID, "rq-1", "", response, updated); err != nil {
t.Fatalf("CompleteReplan happy path: %v", err)
}
got, err := store.GetAttempt(ctx, sessionID)
if err != nil {
t.Fatalf("GetAttempt after replan: %v", err)
}
if got.CurrentReplanRequestID != "rq-1" {
t.Fatalf("current_replan_request_id = %q, want rq-1", got.CurrentReplanRequestID)
}
if got.EffectiveMediaFileID != f.altFileID {
t.Fatalf("effective_media_file_id = %d, want %d", got.EffectiveMediaFileID, f.altFileID)
}
if got.CurrentPlanID != "plan-2" || got.CurrentPlan.PlanID != "plan-2" || got.CurrentPlan.DecisionReason != "transcode_fallback" {
t.Fatalf("plan did not round-trip through replan: %+v", got.CurrentPlan)
}
if !bytes.Equal(mustJSON(t, got.CurrentPlan), mustJSON(t, updated.CurrentPlan)) {
t.Fatalf("plan JSON mismatch after replan:\n got %s\nwant %s", mustJSON(t, got.CurrentPlan), mustJSON(t, updated.CurrentPlan))
}
// The migration's sync trigger must not fight the in-transaction CAS:
// the raw column must equal the new request ID, with no extra rewrite.
var rawRevision, replanState string
if err := f.pool.QueryRow(ctx, `
SELECT a.current_replan_request_id, r.state
FROM playback_v3_attempts a
JOIN playback_v3_replans r ON r.session_id = a.session_id AND r.replan_request_id = $2
WHERE a.session_id = $1::uuid`, sessionID, "rq-1").Scan(&rawRevision, &replanState); err != nil {
t.Fatalf("inspect attempt/replan rows: %v", err)
}
if rawRevision != "rq-1" {
t.Fatalf("raw current_replan_request_id = %q, want rq-1", rawRevision)
}
if replanState != "completed" {
t.Fatalf("replan state = %q, want completed", replanState)
}
// A second replan whose base does not match the current revision must
// lose the compare-and-swap.
if _, err := store.BeginReplan(ctx, sessionID, "rq-2", "rq-digest-2", "rq-1", future); err != nil {
t.Fatalf("BeginReplan second: %v", err)
}
stale := updated
stale.CurrentReplanRequestID = "rq-2"
if err := store.CompleteReplan(ctx, sessionID, "rq-2", "wrong-base", response, stale); !errors.Is(err, playback.ErrReplanSupersededV3) {
t.Fatalf("CompleteReplan wrong base: got %v, want ErrReplanSupersededV3", err)
}
// Unknown session.
if err := store.CompleteReplan(ctx, uuid.NewString(), "rq-1", "", response, updated); !errors.Is(err, playback.ErrSessionNotFound) {
t.Fatalf("CompleteReplan missing session: got %v, want ErrSessionNotFound", err)
}
})
t.Run("CleanupExpired", func(t *testing.T) {
sessionID := uuid.NewString()
attemptID := "att-cleanup-" + sessionID
if err := store.SaveAttempt(ctx, f.attemptRecord(sessionID, attemptID, "digest-cleanup")); err != nil {
t.Fatalf("SaveAttempt: %v", err)
}
if _, err := store.BeginReplan(ctx, sessionID, "rq-1", "rq-digest-1", "", time.Now().Add(time.Minute)); err != nil {
t.Fatalf("BeginReplan: %v", err)
}
// A survivor attempt that must not be swept.
keepSession := uuid.NewString()
keepAttempt := "att-keep-" + keepSession
if err := store.SaveAttempt(ctx, f.attemptRecord(keepSession, keepAttempt, "digest-keep")); err != nil {
t.Fatalf("SaveAttempt survivor: %v", err)
}
event := func(attempt string) playback.RouteEventRecordV3 {
return playback.RouteEventRecordV3{
RouteEventV3: playback.RouteEventV3{
ProtocolVersion: 3,
PlaybackAttemptID: attempt,
SessionID: sessionID,
Event: playback.RouteEventNamesV3()[0],
Diagnostics: map[string]string{},
},
UserID: f.userID,
ProfileID: "profile-1",
}
}
if err := store.RecordRouteEvent(ctx, event("att-cleanup-old")); err != nil {
t.Fatalf("RecordRouteEvent old: %v", err)
}
if err := store.RecordRouteEvent(ctx, event("att-cleanup-recent")); err != nil {
t.Fatalf("RecordRouteEvent recent: %v", err)
}
if _, err := f.pool.Exec(ctx, `
UPDATE playback_route_events SET received_at = NOW() - INTERVAL '31 days'
WHERE playback_attempt_id = 'att-cleanup-old'`); err != nil {
t.Fatalf("age route event: %v", err)
}
f.expireAttempt(t, attemptID)
removed, err := store.CleanupExpired(ctx, time.Now())
if err != nil {
t.Fatalf("CleanupExpired: %v", err)
}
if removed < 1 {
t.Fatalf("CleanupExpired removed %d attempts, want at least 1", removed)
}
var attempts, replans, oldEvents, recentEvents int
if err := f.pool.QueryRow(ctx, `SELECT COUNT(*) FROM playback_v3_attempts WHERE playback_attempt_id = $1`, attemptID).Scan(&attempts); err != nil {
t.Fatalf("count attempts: %v", err)
}
if err := f.pool.QueryRow(ctx, `SELECT COUNT(*) FROM playback_v3_replans WHERE session_id = $1::uuid`, sessionID).Scan(&replans); err != nil {
t.Fatalf("count replans: %v", err)
}
if err := f.pool.QueryRow(ctx, `SELECT COUNT(*) FROM playback_route_events WHERE playback_attempt_id = 'att-cleanup-old'`).Scan(&oldEvents); err != nil {
t.Fatalf("count old events: %v", err)
}
if err := f.pool.QueryRow(ctx, `SELECT COUNT(*) FROM playback_route_events WHERE playback_attempt_id = 'att-cleanup-recent'`).Scan(&recentEvents); err != nil {
t.Fatalf("count recent events: %v", err)
}
if attempts != 0 {
t.Fatalf("expired attempt survived cleanup")
}
if replans != 0 {
t.Fatalf("replans did not cascade with the expired attempt")
}
if oldEvents != 0 {
t.Fatalf("31-day-old route event survived cleanup")
}
if recentEvents != 1 {
t.Fatalf("recent route event count = %d, want 1", recentEvents)
}
if _, err := store.GetAttempt(ctx, keepSession); err != nil {
t.Fatalf("unexpired attempt was swept: %v", err)
}
})
t.Run("AcquireSessionLock", func(t *testing.T) {
sessionID := uuid.NewString()
release1, err := store.AcquireSessionLock(ctx, sessionID)
if err != nil {
t.Fatalf("first AcquireSessionLock: %v", err)
}
// A different session must not be serialized behind the first lock.
otherCtx, cancelOther := context.WithTimeout(ctx, 5*time.Second)
defer cancelOther()
releaseOther, err := store.AcquireSessionLock(otherCtx, uuid.NewString())
if err != nil {
t.Fatalf("different-session AcquireSessionLock blocked: %v", err)
}
releaseOther()
// A second acquire on the same session must block until release.
type lockResult struct {
release func()
err error
}
acquired := make(chan lockResult, 1)
go func() {
release2, err := store.AcquireSessionLock(ctx, sessionID)
acquired <- lockResult{release: release2, err: err}
}()
select {
case result := <-acquired:
if result.err == nil {
result.release()
}
t.Fatalf("second lock acquired while first was held (err=%v)", result.err)
case <-time.After(300 * time.Millisecond):
// Still blocked, as required.
}
release1()
select {
case result := <-acquired:
if result.err != nil {
t.Fatalf("second AcquireSessionLock after release: %v", result.err)
}
result.release()
case <-time.After(5 * time.Second):
t.Fatal("second lock never acquired after first release")
}
// Releasing twice is safe (sync.Once) and the lock is free again.
release1()
release3, err := store.AcquireSessionLock(ctx, sessionID)
if err != nil {
t.Fatalf("reacquire after release: %v", err)
}
release3()
})
}