Two connected failure modes that silently break multi-pod deployments:
1. `RedisURL` has a struct-level default (`redis://<appDomain>:6379`)
that makes `c.RedisURL == ""` always false. An operator forgetting
to set `REDIS_URL` booted against a phantom host — every Redis call
would then fail, and `ChatPubSubService` would quietly fall back to
an in-memory map. On a single-pod deploy that "works"; on two pods
it silently partitions chat (messages on pod A never reach
subscribers on pod B).
2. The fallback itself was logged at `Warn` level, buried under normal
traffic. Operators only noticed when users reported stuck chats.
Changes:
* `config.go` (`ValidateForEnvironment` prod branch): new check that
`os.Getenv("REDIS_URL")` is non-empty. The struct field is left
alone (dev + test still use the default); we inspect the raw env so
the check is "explicitly set" rather than "non-empty after defaults".
* `chat_pubsub.go` `NewChatPubSubService`: if `redisClient == nil`,
emit an `ERROR` at construction time naming the failure mode
("cross-instance messages will be lost"). Same `Warn`→`Error`
promotion for the `Publish` fallback path — runbook-worthy.
Tests: new `chat_pubsub_test.go` with a `zaptest/observer` that asserts
the ERROR-level log fires exactly once when Redis is nil, plus an
in-memory fan-out happy-path so single-pod dev behaviour stays covered.
New `TestValidateForEnvironment_RedisURLRequiredInProduction` mirrors
the Hyperswitch guard test shape.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
172 lines
4 KiB
Go
172 lines
4 KiB
Go
package services
|
|
|
|
import (
|
|
"context"
|
|
"sync"
|
|
|
|
"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 {
|
|
// 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
|
|
}
|
|
|
|
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 {
|
|
return s.redisClient.Publish(ctx, channel, event).Err()
|
|
}
|
|
|
|
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
|
|
}
|