Some checks failed
Veza CI / Backend (Go) (push) Failing after 8m56s
Veza CI / Frontend (Web) (push) Has been cancelled
E2E Playwright / e2e (full) (push) Has been cancelled
Veza CI / Notify on failure (push) Blocked by required conditions
Veza CI / Rust (Stream Server) (push) Successful in 5m3s
Security Scan / Secret Scanning (gitleaks) (push) Failing after 53s
Three Incus containers, each running redis-server + redis-sentinel (co-located). redis-1 = master at first boot, redis-2/3 = replicas. Sentinel quorum=2 of 3 ; failover-timeout=30s satisfies the W3 acceptance criterion. - internal/config/redis_init.go : initRedis branches on REDIS_SENTINEL_ADDRS ; non-empty -> redis.NewFailoverClient with MasterName + SentinelAddrs + SentinelPassword. Empty -> existing single-instance NewClient (dev/local stays parametric). - internal/config/config.go : 3 new fields (RedisSentinelAddrs, RedisSentinelMasterName, RedisSentinelPassword) read from env. parseRedisSentinelAddrs trims+filters CSV. - internal/metrics/cache_hit_rate.go : new RecordCacheHit / Miss counters, labelled by subsystem. Cardinality bounded. - internal/middleware/rate_limiter.go : instrument 3 Eval call sites (DDoS, frontend log throttle, upload throttle). Hit = Redis answered, Miss = error -> in-memory fallback. - internal/services/chat_pubsub.go : instrument Publish + PublishPresence. - internal/websocket/chat/presence_service.go : instrument SetOnline / SetOffline / Heartbeat / GetPresence. redis.Nil counts as a hit (legitimate empty result). - infra/ansible/roles/redis_sentinel/ : install Redis 7 + Sentinel, render redis.conf + sentinel.conf, systemd units. Vault assertion prevents shipping placeholder passwords to staging/prod. - infra/ansible/playbooks/redis_sentinel.yml : provisions the 3 containers + applies common baseline + role. - infra/ansible/inventory/lab.yml : new groups redis_ha + redis_ha_master. - infra/ansible/tests/test_redis_failover.sh : kills the master container, polls Sentinel for the new master, asserts elapsed < 30s. - config/grafana/dashboards/redis-cache-overview.json : 3 hit-rate stats (rate_limiter / chat_pubsub / presence) + ops/s breakdown. - docs/ENV_VARIABLES.md §3 : 3 new REDIS_SENTINEL_* env vars. - veza-backend-api/.env.template : 3 placeholders (empty default). Acceptance (Day 11) : Sentinel failover < 30s ; cache hit-rate dashboard populated. Lab test pending Sentinel deployment. W3 verification gate progress : Redis Sentinel ✓ (this commit), MinIO EC4+2 ⏳ Day 12, CDN ⏳ Day 13, DMCA ⏳ Day 14, embed ⏳ Day 15. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
184 lines
4.3 KiB
Go
184 lines
4.3 KiB
Go
package services
|
|
|
|
import (
|
|
"context"
|
|
"sync"
|
|
|
|
"veza-backend-api/internal/metrics"
|
|
|
|
"github.com/google/uuid"
|
|
"github.com/redis/go-redis/v9"
|
|
"go.uber.org/zap"
|
|
)
|
|
|
|
type ChatPubSubService struct {
|
|
redisClient *redis.Client
|
|
logger *zap.Logger
|
|
inMemorySubscribers map[string][]chan []byte
|
|
mu sync.RWMutex
|
|
}
|
|
|
|
func NewChatPubSubService(redisClient *redis.Client, logger *zap.Logger) *ChatPubSubService {
|
|
if logger == nil {
|
|
logger = zap.NewNop()
|
|
}
|
|
if redisClient == nil {
|
|
// In multi-pod deployments the in-memory fallback silently breaks:
|
|
// messages published on pod A are never seen by subscribers on pod B.
|
|
// Emit a loud startup error so the misconfiguration is noticed.
|
|
logger.Error("Redis unavailable, falling back to in-memory PubSub — cross-instance messages will be lost. Set REDIS_URL and restart for multi-pod correctness")
|
|
}
|
|
return &ChatPubSubService{
|
|
redisClient: redisClient,
|
|
logger: logger,
|
|
inMemorySubscribers: make(map[string][]chan []byte),
|
|
}
|
|
}
|
|
|
|
func (s *ChatPubSubService) roomChannel(roomID uuid.UUID) string {
|
|
return "chat:room:" + roomID.String()
|
|
}
|
|
|
|
func (s *ChatPubSubService) Publish(ctx context.Context, roomID uuid.UUID, message []byte) error {
|
|
channel := s.roomChannel(roomID)
|
|
|
|
if s.redisClient != nil {
|
|
if err := s.redisClient.Publish(ctx, channel, message).Err(); err != nil {
|
|
metrics.RecordCacheMiss("chat_pubsub")
|
|
// ERROR, not Warn: the in-memory fallback only reaches subscribers
|
|
// on this pod — a multi-pod chat becomes partitioned until Redis
|
|
// recovers. Operators should page on this log line.
|
|
s.logger.Error("Redis publish failed, in-memory fallback will not reach other pods",
|
|
zap.String("channel", channel),
|
|
zap.Error(err),
|
|
)
|
|
s.publishInMemory(channel, message)
|
|
return nil
|
|
}
|
|
metrics.RecordCacheHit("chat_pubsub")
|
|
return nil
|
|
}
|
|
|
|
metrics.RecordCacheMiss("chat_pubsub")
|
|
s.publishInMemory(channel, message)
|
|
return nil
|
|
}
|
|
|
|
func (s *ChatPubSubService) Subscribe(ctx context.Context, roomID uuid.UUID) (<-chan []byte, func(), error) {
|
|
channel := s.roomChannel(roomID)
|
|
|
|
if s.redisClient != nil {
|
|
pubsub := s.redisClient.Subscribe(ctx, channel)
|
|
ch := make(chan []byte, 256)
|
|
|
|
go func() {
|
|
defer close(ch)
|
|
for {
|
|
msg, err := pubsub.ReceiveMessage(ctx)
|
|
if err != nil {
|
|
return
|
|
}
|
|
select {
|
|
case ch <- []byte(msg.Payload):
|
|
default:
|
|
s.logger.Warn("PubSub channel full, dropping message", zap.String("channel", channel))
|
|
}
|
|
}
|
|
}()
|
|
|
|
cancel := func() {
|
|
_ = pubsub.Close()
|
|
}
|
|
return ch, cancel, nil
|
|
}
|
|
|
|
return s.subscribeInMemory(channel)
|
|
}
|
|
|
|
func (s *ChatPubSubService) PublishPresence(ctx context.Context, event []byte) error {
|
|
channel := "chat:presence"
|
|
|
|
if s.redisClient != nil {
|
|
if err := s.redisClient.Publish(ctx, channel, event).Err(); err != nil {
|
|
metrics.RecordCacheMiss("chat_pubsub")
|
|
return err
|
|
}
|
|
metrics.RecordCacheHit("chat_pubsub")
|
|
return nil
|
|
}
|
|
|
|
metrics.RecordCacheMiss("chat_pubsub")
|
|
s.publishInMemory(channel, event)
|
|
return nil
|
|
}
|
|
|
|
func (s *ChatPubSubService) SubscribePresence(ctx context.Context) (<-chan []byte, func(), error) {
|
|
channel := "chat:presence"
|
|
|
|
if s.redisClient != nil {
|
|
pubsub := s.redisClient.Subscribe(ctx, channel)
|
|
ch := make(chan []byte, 256)
|
|
|
|
go func() {
|
|
defer close(ch)
|
|
for {
|
|
msg, err := pubsub.ReceiveMessage(ctx)
|
|
if err != nil {
|
|
return
|
|
}
|
|
select {
|
|
case ch <- []byte(msg.Payload):
|
|
default:
|
|
}
|
|
}
|
|
}()
|
|
|
|
cancel := func() {
|
|
_ = pubsub.Close()
|
|
}
|
|
return ch, cancel, nil
|
|
}
|
|
|
|
return s.subscribeInMemory(channel)
|
|
}
|
|
|
|
func (s *ChatPubSubService) publishInMemory(channel string, message []byte) {
|
|
s.mu.RLock()
|
|
defer s.mu.RUnlock()
|
|
|
|
subscribers, ok := s.inMemorySubscribers[channel]
|
|
if !ok {
|
|
return
|
|
}
|
|
|
|
for _, ch := range subscribers {
|
|
select {
|
|
case ch <- message:
|
|
default:
|
|
}
|
|
}
|
|
}
|
|
|
|
func (s *ChatPubSubService) subscribeInMemory(channel string) (<-chan []byte, func(), error) {
|
|
ch := make(chan []byte, 256)
|
|
|
|
s.mu.Lock()
|
|
s.inMemorySubscribers[channel] = append(s.inMemorySubscribers[channel], ch)
|
|
s.mu.Unlock()
|
|
|
|
cancel := func() {
|
|
s.mu.Lock()
|
|
defer s.mu.Unlock()
|
|
|
|
subs := s.inMemorySubscribers[channel]
|
|
for i, sub := range subs {
|
|
if sub == ch {
|
|
s.inMemorySubscribers[channel] = append(subs[:i], subs[i+1:]...)
|
|
break
|
|
}
|
|
}
|
|
close(ch)
|
|
}
|
|
|
|
return ch, cancel, nil
|
|
}
|